Science Applications International Corp. v. United States ( 2021 )

In the United States Court of Federal Claims
SCIENCE APPLICATIONS
INTERNATIONAL CORP.,
Plaintiff,
v.
THE UNITED STATES,
No. 17-cv-825
Defendant,
Filed: August 6, 2021
and
MICROSOFT CORPORATION,
Intervenor-Defendant,
and
L3 TECHNOLOGIES, INC.,
Third-Party Defendant.
Stephen R. Smith, Cooley LLP, Washington, D.C. for Plaintiff. With him on the brief are DeAnna
D. Allen, Erin M. Estevez, Stephen C. Crenshaw, and James P. Hughes, Cooley LLP, Washington,
D.C.; Douglas P. Lobel, Cooley, LLP, Reston, Virginia; and Goutam Patnaik and Gwendolyn
Tawresey, Pepper Hamilton LLP, Washington, D.C.
Alex Hanna, United States Department of Justice, Civil Division, Washington, D.C. for
Defendant. With him on the briefs are Joseph H. Hunt, Assistant Attorney General, Civil
Division, Gary Hausken, Director, Commercial Litigation Branch, Civil Division, and Scott
Bolden, United States Department of Justice, Civil Division, Washington, D.C.
Thomas Lee Halkowski, Fish and Richardson P.C., Washington, D.C., for Intervenor-Defendant.
With him on briefs are Ahmed J. Davis and Kenton W. Freeman, Jr., Fish and Richardson P.C.,
Washington, D.C.; and Proshanto Mukherji, Fish and Richardson P.C., Boston, MA.
William Carl Bergmann, Baker & Hostetler, LLP, Washington, D.C., for Third-Party Defendant.
With him on briefs are Michael Anderson and Cassandra Simmons, Baker & Hostetler, LLP,
Washington, D.C.
1
MEMORANDUM AND ORDER
This patent case involves complex subject matter, important, cutting-edge technology, and
relates to several high-value contracts with the United States totaling billions of dollars. See e.g.,
July 26, 2021 Hearing Transcript (ECF No. 191) at 8:7-17 (referencing $21.9 billion contract
between the Government and Microsoft). 1 In 2017, Plaintiff Science Applications International
Corp. (SAIC), filed its complaint alleging “that the United States infringed the SERVAL Patents
by entering into contracts with Plaintiff's competitors for the manufacture and subsequent use of
night vision goggle weapon systems with specialized heads up displays that allegedly use
Plaintiff's patented technology.” Sci. Applications Int'l Corp. v. United States, 148 Fed. Cl. 268,
269 (2020); Complaint (ECF No. 1) (Compl.) at ¶¶ 2, 37.
The patents at issue are U.S. Patent Nos. 7,787,012 (the ’012 patent), 8,817,103 (the ’103
patent), 9,229,230 (the ’230 patent), and 9,618,752 (the ’752 patent) (collectively, the Soldier
Enhanced Rapid Engagement and Vision in Ambient Lighting or SERVAL 2 patents). See Compl.
¶¶ 1, 3. 3 Collectively, these patents include 91 total claims. See ’012 patent at 9:62-12:3; ’103
1
See also SAIC’s June 1, 2021 Letter: Motion to Compel Government to Produce Documents
under Court of Fed. Cl. R. 26 and 34 (Public Version) (ECF No. 185) at 2 (referencing a March
25, 2021 agreement between the Government and Microsoft under which the Government agreed
to pay $21.9 billion for IVAS HUD devices); Plaintiff Science Applications International Corp.’s
Claim Construction Brief (ECF No. 90) at 5-6 (referencing “contract awards to BAE Systems and
DRS Technologies for not to exceed values of $444.8 million and $367 million, respectively,” “a
May 16, 2018 contract award to L3 Technologies, Inc. worth up to nearly $400 million, and a
November 20, 2018 other transaction agreement with Microsoft worth up to nearly $500
million.”); Defendant’s Objections and Second Supplemental Responses to Plaintiff’s Second Set
of Interrogatories (7-9) (ECF No. 179-2) (SEALED) at 17-29.
2
A serval is “a sub-Saharan African cat that is known to have the best kill rate in the wild cat
family, including when hunting at night.” Pl.’s Technology Tutorial Presentation at 29.
3
The patents are available as attachments to SAIC’s Opening Claim Construction Brief.
Specifically, the ’012 patent is Exhibit A (ECF No. 90-4); the ’103 patent is Exhibit B (ECF No.
90-5); the ’230 patent is Exhibit C (ECF No. 90-6); and the ’752 patent is Exhibit D (ECF No.
90-7).
2
patent at 10:6-59; ’230 patent at 24:24-30:42; ’752 patent at 25:1-28:57. The parties seek
construction on six terms found throughout each of the four asserted patents. This claim
construction Memorandum and Order construes the disputed terms.
BACKGROUND
I. Technology Overview
Night vision goggles (NVGs) passively amplify miniscule amounts of ambient light, such
as starlight, and enable a soldier to see obscured targets in the dark. ’012 patent at 1:34-36.
However, the process of acquiring and striking a target while using these goggles can be
cumbersome. Id. at 1:32-33. When a soldier located a target, the soldier was forced to either (1)
flip the goggles out of the way and reacquire the target with the sight on his weapon or (2) engage
a laser illuminator on his weapon that followed the weapon’s line of sight and thus indicated where
the bullet would strike. Id. at 1:36-48. Both options came with their own drawbacks. If the soldier
employed the first option, the soldier would lose valuable seconds removing his goggles and be
forced to acquire the target using the weapon’s narrower field of vision, which may be virtually
impossible with a distant or moving target. Id. at 1:39-43. If the soldier employed the second
option, the illuminator may have the unintended effect of giving away the soldier’s position
because the laser illuminator may be just as visible to an enemy as it is to the soldier. Id. at
1:44-1:57.
To alleviate these issues associated with acquiring targets using NVGs, U.S. military
planners set out to develop technology that would combine images from a weapon sight and the
NVG’s visual field. See ’012 patent at 1:59-2:27. According to the SERVAL patents, the prior
art combined images from the field of view and the weapon sight but did not “register” the images.
3
Id. at 1:65-2:3. Instead, the patents assert that, under the prior art, the combined images were often
“distinctly offset” with two distinct images of the same target appearing in different places. Id. at
2:13-2:15. This offset could confuse and disorient the soldier because the soldier could have
difficulty determining the location of his weapon sight in relation to his night vision goggle field
of view. Id. at 2:15-2:18.
To resolve these issues found in the prior art, Retired Brigadier General John Scales led a
team of SAIC engineers in developing the technologies claimed in the four asserted patents, which
form two interrelated patent families. See Compl. ¶ 1; Sci. Applications Int’l Corp., 135 Fed. Cl.
at 664. Each patent family is described below in turn.
1. First Patent Family (’012 and ’103 Patents)
The First Patent Family consists of the ’012 and ’103 patents, which share a common
specification. 4 The ’012 patent is entitled “System and Method for Video Image Registration in a
Heads Up Display.” See generally ’012 patent. The ’012 patent was issued on August 31, 2010
from an application filed on December 2, 2004 that did not claim priority to any earlier-filed
application. Id. On its face, the ’012 patent identifies two inventors, John Richard Scales and
Mark David Hose, and an assignee, Science Applications International Corporation of San Diego,
California. Id. The ’012 patent issued with nineteen total claims. See ’012 patent at 9:62-12:3.
Method claims 1 and 17 are the only independent claims. Id. at 9:63-12:3. Claim 1 of the ’012
patent recites:
1. A method of registering video images with an underlying visual field comprising
the steps of:
(1) determining a source orientation of a video source providing a video
feed containing data for a series of video images representing portions of a
visual field;
4
Because the ’012 and ’103 patents share the same specification, this Memorandum and Order
cites to only the ’012 patent when referring to the specification for the First Patent Family.
4
(2) determining a display orientation of a transparent display overlaying the
visual field, wherein the video source and the transparent display are
independently movable about multiple axes; and
(3) displaying the video images in positions on the transparent display that
overlay portions of the visual field represented by the displayed video
images, wherein boundaries of the displayed video images are in
registration with boundaries of portions of the visual field represented by
the displayed video images.
Id. at 9:63-10:10.
Claim 17 of the ’012 patent recites:
17. A method of registering video images with an underlying visual field
comprising the steps of:
(1) determining a source orientation of a video source of a video feed;
(2) determining a display orientation of a transparent display overlaying
the visual field;
(3) displaying a portion of the video feed in the transparent display;
(4) registering the portion of the video feed with the underlying visual
field; and
(5) repositioning the portion of the video feed within the transparent display
when the video source or transparent display moves.
Id. at 10:48-60. Claims 2-16 and 18-19 of the ’012 patent depend ultimately on claim 1. Id. at
9:63- 12:3.
Like the ’012 patent, the ’103 patent is also entitled “System and Method for Video Image
Registration in a Heads Up Display.” See generally ’103 patent. The ’103 patent was issued on
August 26, 2014 from a divisional application filed on July 26, 2010 that claims priority to the
application filed on December 2, 2004, which issued as the ’012 patent. Id. The ’103 patent shares
the same specification with the ’012 patent. Id. On its face, the ’103 patent identifies the same
inventors and assignee as the ’012 patent. Id. The ’103 patent issued with twelve total claims;
system claim 1, reproduced below, is the only independent claim. Id. at 10:7-59.
Claim 1 of the ’103 patent recites:
1. A system comprising:
a video camera adapted to provide, in a video feed, data for
5
a series of video images representing portions of a visual field;
a first orientation sensor adapted to detect an orientation of
the video camera;
a heads up display (HUD) adapted for viewing of the visual
field by a user of the system wherein the HUD comprises
a transparent display, and wherein the HUD and the
video camera are independently movable about multiple
axes;
a second orientation sensor adapted to detect an orientation
of the HUD; and
a computer adapted to receive sensor data from the first
and second orientation sensors, to receive the video feed
from the video camera, and to display the video images,
on the transpar[e]nt display and based on the received
sensor data, in positions that overlay portions of the
visual field represented by the displayed video images
wherein boundaries of the displayed video images are in
registration with boundaries of portions of the visual
field represented by the displayed video images,
and wherein the computer is adapted to
determine a source orientation of the video camera, and
determine a display orientation of the transparent display.
Id. at 10:7-31.
SAIC’s asserted claim 1 of the ’012 patent and asserted claim 1 of the ’103 patent recite
nearly identical steps, though the former is a method and the latter is a system. The First Patent
Family summarizes their purported invention as “a method for aligning video images with an
underlying visual field” by performing various steps. ’012 patent at 2:31-37. Those steps are
“determining a source orientation of a video source, determining a display orientation of a
transparent display overlaying the visual field, and displaying video images in the transparent
display,” where the position of the video “images is based on the source orientation and the display
orientation.” Id. at 2:33-37. In other words, “[a] video camera is coupled with a heads up display,
and a computer positions images from the video camera on the heads up display based on the
relative orientations of the camera and the display.” Id. at Abstract. “The video image, which
may, for example, come from a weapon sight, is aligned within the heads up display. . . .” Id.
6
Figure 5, reproduced below, which appears on the front of the ’012 and ’103 patents, illustrates
the described configuration. ’012 patent at Fig. 5.
The ’012 and ’103 patents admit that prior art methods and systems, including prior art
night vision goggles (such as “Sensor Technology Systems’ Model 2733 Low Profile Night Vision
Goggle”), already “have the ability to port a video feed into a beam combiner, overlaying a video
image from a video source mounted in the weapon sight onto the center of the visual field of the
goggle.” ’012 patent at 1:65-2:3. Figure 1 of the ’012 and ’103 patents illustrate an example of
the combined image generated by the prior art.
7
In Figure 1, the ’012 and ’103 patents describe the problem with the prior art solution as
“the video feed 102 remains stationary in the center of the visual field 101, obscuring content in
the center of the visual field . . . .” ’012 patent at 2:10-13. According to the patents, the combined
images in Figure 1,
depict the same subjects, a group of soldiers accompanying an armored personnel
carrier (APC). However, the video feed 102 remains stationary in the center of the
visual field 101, obscuring content in the center of the visual field, in this case the
APC and a soldier. The two images are distinctly offset, with the two soldiers to
the right of the APC being repeated in both images. This offset, with two distinct
images of the same target appearing in different places in the field of view could
confuse the soldier, causing a delay in engagement or a miss.
’012 patent at 2:9-18.
Under the First Patent Family, images from the weapon sight’s video source and night
vision goggles are “registered” using the relative orientations of the weapon’s sight and the night
vision goggles. Id. at Abstract, 2:4-7. Figure 4 illustrates the combined image using the invention.
8
In Figure 4, the visual field 400 is “the view through a soldier’s night vision goggles or
other (clear) goggles [that] is enhanced with the addition of a portion of the weapon sight video
feed 401 through the use of a heads up display (HUD).” ’012 patent at 3:57-61. “[T]he video feed
401 has been positioned over the portion of the visual field 400 based on the direction the video
source is pointed.” Id. at 3:64-66. “As the weapon moves, the video feed 401 is dynamically
positioned within the visual field 400.” Id. at 3:67-4:1. A side-by-side comparison of Figures 1
and 4 illustrates the advancement over the prior art.
9
The ’012 and ’103 patents provide a single high-level flowchart, in Figure 8, which
“demonstrates an illustrative embodiment of a method for registering a video image with an
underlying visual field.” ’012 patent at 6:25-27.
In Figure 8, the relative orientation of the display and video source is used to determine the
placement (e.g., X and Y position and rotation) of a video frame containing image data on the
display. Id. at 6:25-27, 6:38-7:12. Specifically, Figure 8 explains that the invention registers the
image by: (1) calculating pitch and yaw delta values (804), (2) determining frame location (805),
(3) calculating the roll delta value (806), (4) determining frame rotation (807), and (5) cropping,
resizing, and enhancing the frame (808). Next, “the processed video frame . . . may be displayed
in a heads up display, as in step 809.” Id. at 7:16‐17. As step 810 illustrates, “if another frame of
10
video is set to be received (i.e., the display is still on), then the process repeats for each new frame,
returning to step 802.” Id. at 7:30‐33.
2. Second Patent Family (’230 and ’752 Patents)
The Second Patent Family includes two patents that also share a common specification. 5
The ’230 patent is entitled “System and Method for Video Image Registration and/or Providing
Supplemental Data in a Heads Up Display.” See generally ’230 patent. The ’230 patent was
issued on January 5, 2016 from an application filed on February 28, 2007 that does not claim
priority to any earlier-filed application. Id. On its face, the ’230 patent identifies two inventors,
John Richard Scales and Michael Harris Rodgers, and an assignee, Science Applications
International Corporation of McLean, VA. Id. The ’230 patent issued with forty-two total claims;
claims 1, 15, and 29 are the only independent claims. Id. at 24:24-30:42.
Claim 1 of the ’230 Patent recites:
1. A system, comprising:
a first video source configured to generate images repre-
senting portions of an external environment;
a second video source, movable independent of the first
video source, configured to generate images represent-
ing portions of the external environment;
a video display; and
a controller coupled to the first and second video sources
and to the display, wherein the controller is configured to
(a) receive video images from the first video source and
from the second video source,
(b) receive motion data indicative of motion of the first
and second video sources,
(c) identify, based on the received motion data, a part of
a first video source image that potentially represents a
portion of the external environment represented in a
part of a second video source image;
(d) evaluate, based on a comparison of data from the first
and second video source images, the identification
performed in operation (c); and
5
Because the ’230 and ’752 patents share the same specification, this Memorandum and Order
cites to only the ’230 patent when referring to the specification for the Second Patent Family.
11
(e) display at least a portion of the first video source
image and at least a portion of the second video source
image such that the second video source image por-
tion overlays a corresponding region of the first video
source image portion, wherein the corresponding
region represents a portion of the external environ-
ment represented in the second video source portion.
Id. at 24:25-51.
Claims 2-14 of the ’230 patent depend ultimately on claim 1. Id. at 24:51-26:26.
Similarly, claim 15 of the ’230 patent mirrors claim 1, except that it is a method claim
rather than a system claim, and it recites:
15. A method, comprising:
(a) receiving video images from a first video source and
from a second video source representing portions of an
external environment;
(b) receiving motion data indicative of motion of the first
and second video sources;
(c) identifying, based on the received motion data, a part of
a first video source image that potentially represents a
portion of the external environment represented in a part
of a second video source image;
(d) evaluating, based on a comparison of data from the first
and second video source images, the identification per-
formed in step (c); and
(e) displaying at least a portion of the first video source
image and at least a portion of the second video source
image such that the second video source image portion
overlays a corresponding region of the first video source
image portion, wherein the corresponding region repre-
sents a portion of the external environment represented
in the second video source portion.
Id. at 26:27-47.
Claims 16-28 of the ’230 patent depend ultimately on claim 15. Id. at 26:48-28:15.
Claim 29 of the ’230 patent recites the same steps as claims 1 and 15, except it involves a
“non-transitory machine-readable medium having machine-executable instructions for performing
a method.” It recites in full:
12
29. A non-transitory machine-readable medium having
machine-executable instructions for performing a method,
comprising:
(a) receiving video images from a first video source
and from a second video source representing portions of an
external environment;
(b) receiving motion data indicative of motion of the first
and second video sources;
(c) identifying, based on the received motion data, a part of
a first video source image that potentially represents a
portion of the external environment represented in a part
of a second video source image;
(d) evaluating, based on a comparison of data from the first
and second video source images, the identification per-
formed in step (c); and
(e) displaying at least a portion of the first video source
image and at least a portion of the second video source
image such that the second video source image portion
overlays a corresponding region of the first video source
image portion, wherein the corresponding region repre-
sents a portion of the external environment represented
in the second video source portion.
Id. at 28:17-38.
Claims 30-42 of the ’230 patent depend ultimately on claim 29. Id. at 28:39-30:42.
Like the ’230 patent, the ’752 patent is also entitled “System and Method for Video Image
Registration and/or Providing Supplemental Data in a Heads Up Display.” See generally ’752
patent. The ’752 patent was issued on April 11, 2017 from a continuation application filed on
November 24, 2015 that claims priority to the application filed on February 28, 2007, which issued
as the ’230 patent. Id. On its face, the ’752 patent identifies the same inventors and assignee as
the ’230 patent. Id. The ’752 patent issued with eighteen total claims; claims 1, 7, and 13, all
reproduced below, are the only independent claims. Id. at 25:2-28:58. Claim 1 of the ’752 patent
is a system claim and recites in full:
1. A system, comprising:
a first video source configured to generate video data
of images representing portions of an external environ-
13
ment within a field of view of the first video source;
a second video source configured to generate second
video data of images representing portions of the
external environment within a field of view of the
second video source;
a video display; and
a controller in communication with the first and second
video sources and the display and configured to per-
form operations that include
receiving the first video data, the second video data,
first motion data corresponding to the first video
source, and second motion data corresponding to the
second video source,
identifying, based on the received first motion data and
the received second motion data, a region of a first
image generable from the first video data for com-
parison with a region of a second image generable
from the second video data,
comparing data corresponding to the identified region
of the first image and data corresponding to the
region of the second image,
selecting, based on the comparing, a part of the first
image and a part of the second image that represent
the same portion of the external environment,
displaying at least a portion of the first image and the
selected part of the second image such that the
selected part of the second image replaces the
selected part of the first image and is in registration
with regions of the first image surrounding the
selected part of the first image.
Id. at 25:2-34. Claims 2-6 of the ’752 patent depend ultimately on claim 1. Id. at 25:35-26:19.
Claim 7 of the ’752 patent is a method claim and recites:
7. A method comprising:
receiving first video data of images representing portions
of an external environment within a field of view of a
first video source;
receiving second video data of images representing por-
tions of the external environment within a field of view
of a second video source;
receiving first motion data corresponding to the first video
source and second motion data corresponding to the
second video source;
identifying, based on the received first motion data and
14
the received second motion data, a region of a first
image generable from the first video data for compar-
ison with a region of a second image generable from the
second video data;
comparing data corresponding to the identified region of
the first image and data corresponding to the region of
the second image;
selecting, based on the comparing, a part of the first image
and a part of the second image that represent a same
portion of the external environment; and
displaying at least a portion of the first image and the
selected part of the second image such that the selected
part of the second image replaces the selected part of
the first image and is in registration with regions of the
first image surrounding the selected part of the first
image.
Id. at 26:19-26:45. Claims 8-12 of the ’752 patent depend ultimately on claim 7. Id. at
26:48-27:30.
Claim 13 of the ’752 patent involves a “non-transitory machine-readable medium
having machine executable instructions for performing a method” and recites in full:
13. A non-transitory machine-readable medium having
machine executable instructions for performing a method
comprising:
receiving first video data of images representing portions
of an external environment within a field of view of a
first video source;
receiving second video data of images representing por-
tions of the external environment within a field of view
of a second video source;
receiving first motion data corresponding to the first video
source and second motion data corresponding to the
second video source;
identifying, based on the received first motion data
and the received second motion data, a region of a first
image generable from the first video data for compari-
son with a region of a second image generable from the
second video data;
comparing data corresponding to the identified region of
the first image and data corresponding to the region of
the second image;
15
selecting, based on the comparing, a part of the first image
and a part of the second image that represent a same
portion of the external environment; and
displaying at least a portion of the first image and the
selected part of the second image such that the selected
part of the second image replaces the selected part of
the first image and is in registration with regions of the
first image surrounding the selected part of the first
image.
Id. at 27:31-59.
The ’230 and ’752 patents summarize their purported invention as “a computer receives
images from two video sources[, where] [e]ach of those two video sources is movable independent
of the other and generates images that represent a portion of an external environment within its
field of view.” ’230 patent at 1:58-62. “Sensors coupled to the two video sources provide data to
the computer that indicates the spatial orientations of those sources.” Id. at 1:64-66. “Using the
sensor data, the computer determines a location for placing a video image (or a portion thereof)
from a second of the sources (e.g., a rifle-mounted source) in the video image from a first of the
sources (e.g., a goggles-mounted source).” Id. at 1:66-2:3. After a location is determined from
the sensor data, “the two images are displayed such that the second source image (or a portion of
that image) overlays a corresponding portion of the first source image.” Id. at 2:11-14.
The Second Patent Family incorporates by reference the ’012 patent. Id. at 1:19-22. The
’230 and ’752 patents admit that prior art methods and systems include the ’012 patent. Id. at
1:17-34. An important difference between the ’012 patent family and the later ’230 patent family
is the manner in which the images from the weapon sight (first video source) and the night vision
goggles (second video source) are aligned. Markman Hearing Tr. (ECF No. 159) at 23:20-25:7.
The ’012 patent family claims a system in which the images from two different sources are aligned
using only orientation data. See, e.g., ’012 patent, claims 1, 17; ’103 patent, claim 1. By the time
16
the ’230 patent was filed three years later, however, the named inventors had realized that use of
orientation data alone may pose problems. ’230 patent at 1:35-43 (identifying disadvantages of a
system using only sensor data to match images). “For example, many low-cost [inertial
measurement unit (IMU)] sensors experience bias drift over time” that “can result in relative
orientation errors of several degrees per hour.” ’230 patent at 1:38-41. These errors require the
user to periodically recalibrate the IMU sensors, and thus “can disrupt system operation.” Id. at
1:42-44. Thus, the later filed ’230 patent describes an improved, two-step alignment method. The
’230 patent family first uses data from motion sensors to help align images from two different
sources. Id. at 2:6-17. The ’230 patent then performs a second step of comparing the content of
the images themselves to evaluate whether the alignment is correct and to adjust the alignment as
necessary. Id. at 2:6-17, Abstract (“The sensor-based location is checked (and possibly adjusted)
based on a comparison of the images.”). The purported invention of the ’230 and ’752 patents
apparently minimizes the need for such manually initiated recalibration. Id. at 1:44-45, 2:14-16.
Figure 4 of the ’230 patent is an example of a displayed image resulting from this two-step
alignment process.
17
To achieve the image in Figure 4, a computer receives video images from the weapon‐ and
goggles‐mounted sources and inertial data from the sensors. Id. at 6:54-60. The computer
calculates a location for an image from the weapon‐mounted source within an image from the
goggles‐mounted source using the inertial sensor data. Id. at 7:19-22. The sensor-based location
is checked (and possibly adjusted) based on a comparison of the images. Id. at Abstract. “Figure
4 shows an example of a user display 70 provided by the goggles 11.” Id. at 6:54-55. “Located
within the goggles[’] [field of view] (and thus in [the] goggles image 82) are numerous trees and
bushes . . . as well as soldiers 71 (partially behind a tree in the lower left) and 72 (partially covered
by foliage in the upper right).” Id. at 6:62-67. “The [heads up display] portion of [the] user display
70 is shown as a rectangular region 73 in the center portion of the goggles[’] [field of view].” Id.
at 7:3-5. “[O]verlaid on [the heads up display] 73 is a weapon view 74 corresponding to (and
generated from) the scope image.” Id. at 7:6-8. “[T]he location and rotation of [the] weapon view
74 within [the] user display 70 is determined by [the] computer 30 based on output from [the]
sensors 13 and 18 and based on [a] comparison of the scope image with the goggles image.” Id.
18
at 7:19-22. “As [the] rifle 19 is moved, scope images (or portions thereof) are dynamically
positioned within [the] user display 70 so as to indicate where [the] scope 17 (and thus [the] rifle
19) is pointing.” Id. at 7:22-25.
The ’230 and ’752 patents provide a high-level flowchart in Figures 5A-5B, demonstrating
the Second Patent Family’s advancement. Id. at 2:48-49. Figures 5A-5B, reproduced below, “are
a flow chart explaining the operation of [the] system 10.” Id. at 7:46-47. As shown below, initial
calibration occurs at step 103 and recalibration, if necessary, occurs at step 117 “thereby correcting
for bias drift and helping to maintain proper registration of the scope image within the goggles
image.” Id. at 7:61-64, 10:4-15.
19
II. Procedural History
Familiarity with the background of this litigation is presumed. See, e.g., Sci. Applications
Int’l Corp. v. United States, No. 17-CV-825, 2021 WL 1568815, at *1 (Fed. Cl. Apr. 21, 2021)
(denying Plaintiff’s motion to strike Defendant’s indefiniteness contentions but awarding fees for
untimely assertion of indefiniteness contentions); Sci. Applications Int’l Corp. v. United States,
20
148 Fed. Cl. 268 (2020) (permitting Government’s Rule 14(b) notice); Sci. Applications Int’l Corp.
v. United States, 135 Fed. Cl. 661, 662 (2018) (denying Government’s motion to dismiss). For
ease of reference, the Court summarizes the litigation as follows.
On June 19, 2017, SAIC filed suit seeking compensation for the Government’s alleged
infringement of the asserted patents. See generally Compl. This suit was based on the United
States’ contract awards to BAE Systems and DRS Technologies for the development of technology
which included the implementation of a Rapid Target Acquisition (RTA) feature relevant to
Plaintiff's infringement claims. Compl. ¶¶ 3, 38.
On November 20, 2018, Microsoft entered into a contract with the United States to develop
an Integrated Visual Augmentation System (IVAS), which Microsoft alleges includes
implementation of an RTA feature relevant to Plaintiff's infringement claims. See Microsoft
Motion to Intervene (ECF No. 59) at 1. Subsequently, Microsoft moved to intervene in this action
“for the limited purpose of protecting its interests regarding the United States’ defense that
products incorporating [RTA] do not infringe the patents asserted in this matter by [Plaintiff].”
Microsoft Motion to Intervene (ECF No. 59) at 1. On May 6, 2019, the Court granted Defendant-
Intervenor Microsoft Corporation's unopposed motion to intervene in this action. See Order
Granting Intervention (ECF No. 60).
On May 30, 2019, nearly two years after the inception of this suit, the Army entered into
two separate other transaction agreements (OTAs) with L3 and Harris, each of which had a night
vision technology division, to develop a prototype for an Enhanced Night Vision Goggle-
Binocular (ENVG-B). Sci. Applications Int'l Corp., 148 Fed. Cl. at 269-70. 6
6
“L3 and Harris merged on June 29, 2019, approximately one month after the L3/Harris contracts
were executed, creating a new entity known as L3Harris Technologies, Inc. . . . L3’s legacy night
vision division remained with the merged entity, and is now known as the Integrated Vision
21
On July 18, 2019, the period for claim construction discovery closed, and the parties jointly
filed deposition transcripts of their respective claim construction experts. See Science
Applications International Corporation’s and the United States’ Joint Submission of Claim
Construction Experts’ Deposition Testimony (ECF No. 79). On October 15, 2019, claim
construction briefs were completed. See Defendant-Intervenor Microsoft Corporation’s Opening
Claim Construction Brief (ECF No. 87) (Microsoft’s Opening Cl. Constr. Br.); Defendant Opening
Claim Construction Brief (ECF No. 89) (Government’s Opening Cl. Constr. Br.); Plaintiff Science
Applications International Corp.’s Claim Construction Brief (ECF No. 90) (Pl.’s Opening Cl.
Constr. Br.); Defendant-Intervenor Microsoft Corporation’s Responsive Claim Construction Brief
(ECF No. 95) (Microsoft’s Responsive Cl. Constr. Br.); Plaintiff Science Applications
International Corp.’s Responsive Claim Construction Brief (ECF No. 96) (Pl.’s Responsive Cl.
Constr. Br.); Defendant United States’ Responsive Claim Construction Brief (ECF No. 97)
(Government’s Responsive Cl. Constr. Br.).
After the parties completed claim construction briefing, Microsoft Corporation moved to
stay this action pending a decision from the United States Patent Trial and Appeal Board (PTAB)
concerning whether to institute inter partes review (IPR) of the patents-in-suit. See Microsoft’s
Motion to Stay Proceedings (ECF No. 92). On November 26, 2019, the prior judge overseeing
this case, the Honorable Mary Ellen Coster Williams, granted a partial stay pending a decision by
the PTAB on whether to institute inter partes review of the patents-in-suit. See Order for Partial
Stay (ECF No. 102). On January 27, 2020, the PTAB issued a decision denying institution on
Solutions Sector of L3 Technologies, Inc., which is a subsidiary of L3Harris Technologies, Inc.
. . . Former Harris Corporation’s legacy night vision technology division, which ultimately
received one of the Army's two separate May 30, 2019 OTAs, was spun-off (for regulatory
reasons) and purchased by Elbit Systems of America, LLC (‘Elbit’), which is the U.S. subsidiary
of Elbit Systems, Ltd.” Sci. Applications Int’l Corp., 148 Fed. Cl. at 270 n.1 (internal citation and
quotation omitted).
22
each of Microsoft’s five petitions for IPR. Microsoft Corp. v. Sci. Applications Int’l Corp., No.
IPR2019-01311, 2020 WL 572706, at *1 (P.T.A.B. Jan. 27, 2020).
“In November 2019, during the pendency of the stay, Defendant's counsel first learned
from the Army of the May 2019 L3/Harris contracts, and that the prototypes required by the
contracts may implicate the accused technology in this action.” Sci. Applications Int'l Corp., 148
Fed. Cl. at 270 (internal citations omitted). “On February 20, 2020, the Court lifted the stay, and
six days later, on February 26, 2020, the Government notified SAIC that the Government intended
to move to notify L3 and Harris pursuant to Rule 14(b).” Id.
On February 27, 2020, the Clerk of Court transferred this case to the undersigned judge.
See Order Reassigning Case (ECF No. 112). On March 10, 2020, the Government moved for
leave to notice L3 Technologies, Inc. and Harris Corporation under Rule 14(b) to potentially
appear as third parties to this lawsuit; the Court granted the motion on May 12, 2020. See Sci.
Applications Int’l Corp., 148 Fed. Cl. at 273. That same day, this Court held a status conference
during which it established a schedule for the proceedings in this case through claim construction.
See May 12, 2020 Scheduling Order (ECF No. 121).
After the status conference, L3 intervened in this case. See June 17, 2020 Order (ECF No.
128); L3’s Answer (ECF No. 131). 7 Subsequently, the Court permitted L3 to file a claim
construction brief and permitted SAIC to respond. See September 4, 2020 Scheduling Order (ECF
No. 141). In its opening claim construction brief, L3 proposed additional disputed terms and a
new term construction. See L3’s Opening Claim Construction Brief (ECF No. 148) (L3’s Opening
Cl. Constr. Br.).
7
Elbit Systems of America, LLC declined to intervene in this action. (ECF No. 135).
23
On November 20, 2020, SAIC filed its responsive supplemental claim construction brief
addressing L3’s arguments. See Plaintiff Science Applications International Corp.’s Responsive
Claim Construction Brief to Third-party Defendant L3 Technolgies, [sic] Inc. (ECF No. 151) (Pl.’s
Responsive Cl. Constr. Br. to L3) at 15-20.
At the conclusion of the parties’ briefing, they had requested construction on the following
ten (10) disputed terms or groups of terms:
1. “video camera;” “video source”
2. “external environment”
3. “video images” / “video source image” / “video data of images”
4. “orientation”
5. “motion data”
6. “based on a comparison”
7. “transparent display” / “HUD comprising a transparent display”
8. “overlay terms”
9. “registration” / “registering terms”; “Wherein the boundaries are in
registration/registering the portion of the video feed with the underlying visual field”
10. “twist rotation”
See November 30, 2020 Joint Status Report (ECF No. 153) at 2-3.
On December 1, 2020, the Court held a pre-Markman Hearing status conference during
which the Court encouraged the parties to narrow their disputes where possible. Pre-Markman
Hearing Tr. at 21:1-23:25. Accordingly, on December 10, 2020, parties limited the terms in
dispute to the following:
24
1. “video images” / “video source image” / “video data of images”
2. “transparent display” / “HUD comprising a transparent display”
3. “overlay terms”
4. “based on a comparison”
5. “motion data”
6. “registration” / “registering terms.”
See December 10, 2020 Joint Status Report (ECF No. 157) at 2.
The Court held a Markman Hearing on December 15, 2020, which included the parties’
technology tutorials and arguments on (1) “video images” / “video source image” / “video data of
images,” (2) “transparent display” / “HUD comprising a transparent display,” (3) “motion data,”
(4) “registration” / “registering terms.” See generally Markman Hearing Tr. The parties agreed
to rely on their briefs for the “overlay” and “based on a comparison” terms. See January 8, 2021
Joint Status Report (ECF No. 161).
APPLICABLE LEGAL STANDARDS
I. Claim Construction
“It is a bedrock principle of patent law that the claims of a patent define the invention to
which the patentee is entitled the right to exclude.” Phillips v. AWH Corp., 415 F.3d 1303, 1312
(Fed. Cir. 2005) (en banc) (internal quotations and citations omitted). Claim construction is the
process of determining the meaning and scope of patent claims. Markman v. Westview
Instruments, Inc., 52 F.3d 967, 976 (Fed. Cir. 1995) (en banc), aff’d, 517 U.S. 370 (1996). “[T]he
words of a claim are generally given their ordinary and customary meaning,” which is “the
meaning that the term would have to a person of ordinary skill in the art in question at the time of
25
the invention, i.e., as of the effective filing date of the patent application.” Phillips., 415 F.3d at
1312-13 (internal quotations and citations omitted). 8
The analysis of any disputed claim terms begins with the intrinsic evidence of record, as
“intrinsic evidence is the most significant source of the legally operative meaning of disputed claim
language.” Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582 (Fed. Cir. 1996). “[T]he
claims themselves provide substantial guidance as to the meaning of particular claim terms.”
Phillips., 415 F.3d at 1314 (internal quotation and citations omitted).
“[T]he person of ordinary skill in the art is deemed to read the claim term not only in the
context of the particular claim in which the disputed term appears, but in the context of the entire
8
The Government submits that the level of skill of a person of ordinary skill in the art for the First
and Second Patent Family is a Bachelor of Science degree in Computer Science and 3 years of
experience in programming video/graphics applications and computer vision, or alternatively, a
Master of Science degree in Computer Science and 1 year of experience in programming
video/graphics applications and computer vision. Neumann Decl. (ECF No. 66) ¶ 48. L3 agrees
with the Government. Def. L3 Opening Cl. Constr. Br. at 2 n.1.
“SAIC submits that a person having ordinary skill in the art for the Asserted Patents possesses
either: (1) a Bachelor’s degree in computer science, computer engineering, electrical engineering,
or systems engineering and at least 2 years of experience working with sensor (e.g., camera and
orientation sensor) technology; or (2) a Master’s degree in one of these disciplines and at least 1
year of work experience with the aforementioned sensor technology.” Pl.’s Responsive Cl. Constr.
Br. at 2 (citing Ex. L, Declaration of Dr. Gregory F. Welch ¶ 34)). SAIC notes that its definition
“differs from the Government’s (which requires a degree in Computer Science, excludes persons
with engineering degrees, and focuses on experience in video graphics and computer vision rather
on than on the camera and orientation sensor technology underlying the inventions).” Id. (internal
citations omitted).
At the Markman Hearing, the parties agreed that the dispute regarding a person of ordinary skill
in the art (POSITA) did not need to be decided to resolve the parties’ disputes over Claim
Construction. Markman Hearing Tr. at 60:16-61:7; see also Pl.’s Responsive Cl. Constr. Br. at 2.
Because the parties have indicated that their dispute over the level of ordinary skill in the art does
not affect this Court’s construction of the disputed terms, this Court declines to articulate the level
of a POSITA at this stage in the proceedings. In re Fought, 941 F.3d 1175, 1179 (Fed. Cir. 2019)
(“Unless the patentee places the level of ordinary skill in the art in dispute and explains with
particularity how the dispute would alter the outcome, neither the Board nor the examiner need
articulate the level of ordinary skill in the art.”).
26
patent, including the specification.” Id. at 1313. “[T]he specification ‘is always highly relevant
to the claim construction analysis. Usually, it is dispositive; it is the single best guide to the
meaning of a disputed term.’” Id. at 1315 (quoting Vitronics, 90 F.3d at 1582).
Notwithstanding the importance of a specification, limitations in the specification must not
be read into the claims absent lexicography or disclaimer/disavowal. Hill-Rom Services, Inc. v.
Stryker Corp., 755 F.3d 1367, 1371 (Fed. Cir. 2014). The United States Court of Appeals for the
Federal Circuit (Federal Circuit) has “expressly rejected the contention that if a patent describes
only a single embodiment, the claims of the patent must be construed as being limited to that
embodiment.” Phillips, 415 F.3d at 1323 (internal citations omitted). Conversely, “an
interpretation [which excludes a preferred embodiment] is rarely, if ever, correct and would require
highly persuasive evidentiary support.” Vitronics, 90 F.3d at 1583.
The prosecution history of a patent is also relevant intrinsic evidence. Markman, 52 F.3d
at 980. Although “the prosecution history represents an ongoing negotiation between the PTO and
the applicant, rather than the final product of that negotiation” and for this reason “often lacks the
clarity of the specification,” the prosecution history can nonetheless “often inform the meaning of
the claim language by demonstrating how the inventor understood the invention and whether the
inventor limited the invention in the course of prosecution, making the claim scope narrower than
it would otherwise be.” Phillips, 415 F.3d at 1317 (citations omitted). “[A] patentee’s statements
during prosecution, whether relied on by the examiner or not, are relevant to claim interpretation.”
Microsoft Corp. v. Multi-Tech Sys., Inc., 357 F.3d 1340, 1350 (Fed. Cir. 2004).
“Although [the Federal Circuit has] emphasized the importance of intrinsic evidence in
claim construction, [it has] also authorized district courts to rely on extrinsic evidence, which
‘consists of all evidence external to the patent and prosecution history, including expert and
27
inventor testimony, dictionaries, and learned treatises.’” Phillips, 415 F.3d at 1317 (citing
Markman, 52 F.3d at 1583). While sometimes helpful, extrinsic evidence is “less significant than
the intrinsic record in determining the legally operative meaning of claim language.” Id. at 1317
(quoting C.R. Bard, Inc. v. U.S. Surgical Corp., 388 F.3d 858, 862 (Fed Cir. 2004)).
II. Definiteness
The Patent Act of 1952—which preceded the America Invents Act and, which the parties
agree, is applicable to the patents at issue in this case—requires that “a specification ‘conclude
with one or more claims particularly pointing out and distinctly claiming the subject matter which
the applicant regards as his invention.’” Nautilus, Inc. v. Biosig Instruments, Inc., 572 U.S. 898,
901 (2014) (quoting 35 U.S.C. § 112, ¶ 2 (2006 ed.) (emphasis omitted)). 9 The question of whether
a claim is sufficiently definite under 35 U.S.C. § 112, ¶ 2, is closely related to the claim
construction task because the decision depends on the ability of the court to interpret a claim. “If
a claim is indefinite, the claim, by definition, cannot be construed.” Enzo Biochem, Inc. v. Applera
Corp., 599 F.3d 1325, 1332 (Fed. Cir. 2010). A claim fails to satisfy this statutory requirement
and is thus invalid for indefiniteness if its language, when read in light of the specification and the
prosecution history, “fail[s] to inform, with reasonable certainty, those skilled in the art about the
scope of the invention.” Nautilus, 572 U.S. at 901. However, “the certainty which the law requires
in patents is not greater than is reasonable, having regard to their subject-matter.” Id. at 910 (citing
Minerals Separation, Ltd. v. Hyde, 242 U.S. 261, 270 (1916)). The definiteness standard must
allow for “some modicum of uncertainty” to provide incentives for innovation but must also
9
Paragraph 2 of 35 U.S.C. § 112 was replaced with newly designated § 112(b) when § 4(c) of the
America Invents Act (AIA), Pub. L. No. 112–29, took effect on September 16, 2012. Because the
applications resulting in the patents at issue in this case were filed before that date, this Court will
refer to the pre-AIA version of § 112. Markman Hearing Tr. at 98:9-15.
28
require “clear notice of what is claimed, thereby apprising the public of what is still open to them.”
Id. at 909 (internal quotation marks and citations omitted). It also serves as a “meaningful . . .
check” against “foster[ing] [an] innovation-discouraging ‘zone of uncertainty.’” Id. at 910-11
(quoting United Carbon Co. v. Binney & Smith Co., 317 U.S. 228, 236 (1942)).
Issued patents are presumed valid, and indefiniteness must be proven by clear and
convincing evidence. See Teva Pharm. USA, Inc. v. Sandoz, Inc., 789 F.3d 1335, 1345 (Fed. Cir.
2015); Microsoft Corp. v. i4i Ltd., P’ship, 564 U.S. 91, 95 (2011); see also 35 U.S.C. § 282(a)
(“The burden of establishing invalidity of a patent or any claim thereof shall rest on the party
asserting such invalidity.”).
29
DISCUSSION
I. “VIDEO IMAGES” / “VIDEO SOURCE IMAGE” / “VIDEO DATA OF IMAGES”
SAIC 10 Government Microsoft L3
“Plain and ordinary digital or analog digital or analog digital or analog
meaning (e.g., video frames video frames video frames
electronic images and
data)”
Court’s Construction: “digital, analog, or nonstandard video frames”
Microsoft argues that the patents’ specifications equate “video images” to “video frames,”
and accordingly urges the Court to construe “video image” as “digital or analog video frames.”
See Defendant-Intervenor Microsoft Corporation’s Opening Claim Construction Brief
(Microsoft’s Opening Br.) (ECF No. 87) at 3-4; Defendant-Intervenor Microsoft Corporation’s
Responsive Claim Construction Brief (Microsoft’s Responsive Br.) (ECF 95) at 1-2; Markman
Hearing Tr. at 61:8-23. This proposed construction is, according to Microsoft, the plain and
ordinary meaning of the terms as dictated by the patent specification. See Microsoft’s Opening
Br. at 3; Microsoft’s Responsive Br. at 1-2. To support its proposed construction, Microsoft
highlights that the First Patent Family uses “frame” or “video frame” interchangeably with “video
10
In its briefs, SAIC argued that the terms “‘video images’/‘video source image’ mean
[Images][An image] generated from visual information (e.g., visible light, ambient light,
thermal/IR data, etc.) captured from a video source, depicting an external area or region in the
video source’s field of view.” Pl. Opening Cl. Constr. Br. at 10 (brackets in original). SAIC’s
construction changed during the Markman Hearing; counsel for SAIC argued instead that the better
construction was the term’s plain and ordinary meaning. Markman Hearing Tr. at 51:20-52: 18.
“If there's a construction -- which we don't even think there should be -- but if there is, it should
be, just very simply, electronic images or electronic representations.” Markman Hearing Tr. at
55:13-16. Plaintiff also stated that it would not object to defining “video images” as “digital or
analog or nonstandard images and data.” Markman Hearing Tr. at 88:2-12.
30
image.” See Microsoft’s Opening Br. at 4; Markman Hearing Tr. at 65:3-66:9. More concisely
stated, Microsoft maintains that “frames are literally what defines what video is” regardless of
“[w]hether it’s analog or digital, [video] consists of a series of frames.” Markman Hearing Tr. at
65:22-24.
The Government and L3 agreed with and adopted Microsoft’s proposed construction. See
Government Opening Cl. Constr. Br. at 42-43; Government’s Responsive Cl. Constr. Br. at 1; L3’s
Opening Cl. Constr. Br. at 17-18. At the Markman Hearing, L3 added that the Defendants’
proposed construction is further supported by the Second Patent Family’s use of the term “i.e.,”
when discussing the two terms which, according to L3, indicates that the inventors equated the
terms “image” with “frame.” See Markman Hearing Tr. at 78:4-79:13.
SAIC contended for the first time during the Markman Hearing that the “video images”
claim term should be construed as its plain and ordinary meaning, which SAIC states is “just
electronic images or data.” Markman Hearing Tr. at 51:20-53:5-19, 76:22-77:12. This
construction differed from the construction SAIC offered in its briefs, which proposed construing
the video image terms as “[Images][An image] generated from visual information (e.g., visible
light, ambient light, thermal/IR data, etc.) captured from a video source, depicting an external area
or region in the video source’s field of view.” See Pl.’s Claim Constr. Br. at 10; Pl.’s Responsive
Claim Constr. Br. at 2-3 (brackets in original); Pl.’s Responsive Cl. Constr. Br. to L3 at 12. During
the Markman Hearing, SAIC’s counsel explained this shift in proposed construction by stating that
both SAIC’s original construction and their current construction are intended to combat a format
limitation for video images. Markman Hearing Tr. at 52:10-55:20.
At the Markman Hearing, both parties made important concessions. First, SAIC
acknowledged that its proposed construction does not include optical images. Markman Hearing
31
Tr. at 58:11-25. Second, Microsoft indicated that it would be satisfied with construing images as
digital, analog, or nonstandard video frames to alleviate any of Plaintiff’s concerns related to
limiting video images to “digital or analog formats.” See Markman Hearing Tr. at 72:17-22. Thus,
at this stage in the proceedings, the primary difference between the parties’ proposed constructions
is Defendants’ use of “frames,” which Plaintiff argues improperly limits the scope of the term
“video images.” Markman Hearing Tr. at 58:4-10.
The Court agrees with Defendants’ construction. The terms “video images” / “video source
image” / “video data of images” appear in claims 1, 2, 6, 8, 9, 12, 14, 17, and 19 of the ’012 patent;
claims 1, 4, and 9 of the ’103 patent; claims 1, 3, 4, 5, 7, 9, 10, 13, 15, 17, 18, 19, 21, 23, 24, 27,
29, 31, 32, 33, 35, 37, 38, and 41 of the ’230 Patent; and claims 1, 2, 7, 8, 13, and 14 of the ’752
patent. The parties appear to agree that the terms “video images” / “video source image” / “video
data of images” should be construed consistently throughout both patent families. Pl.’s Opening
Cl. Constr. Br. at 10, 13; Microsoft’s Opening Br. at 3; Government’s Opening Cl. Constr. Br. at
42-43; L3’s Opening Cl. Constr. Br. at 17-18.
While the patents do not define the “video image” terms explicitly, the specifications do
so implicitly through consistent use of comparative language to discuss image data from the first
and second video sources. See Irdeto Access, Inc. v. EchoStar Satellite Corp., 383 F.3d 1295,
1300 (Fed. Cir. 2004) (“Even when guidance is not provided in explicit definitional format, ‘the
specification may define claim terms by implication such that the meaning may be found in or
ascertained by a reading of the patent documents.’” (quoting Bell Atl. Network Servs., Inc. v. Covad
Communications Group, Inc., 262 F.3d 1258, 1268 (Fed. Cir. 2001)); Vitronics, 90 F.3d at 1582(“The specification acts as a dictionary when it expressly defines terms used in the claims or when
it defines terms by implication.”).
32
Throughout both patent families, the phrase “video images” is consistently equated with
still video frames. The specification’s description of Figure 8 of the’012 patent is illustrative.
Figure 8 of the’012 patent, “demonstrates . . . a method for registering a video image with an
underlying visual field.” ’012 patent at 6:25-27. The specification explains in detail how this
image is received, processed, and displayed as a video frame. Id. at 6:25-33. The process for
registration begins “at step 802, a video frame is received for processing. The frame may be
processed digitally, and if it is received in analog form may first need to be converted to a digital
format for processing.” Id. at 6:33-37. At the processing step, “the location of the processed frame
within a heads up display is determined, as in step 805.” Id. at 6:50‐51. Before displaying in the
HUD, “the processed frame can be rotated for presentation within the heads up display, as in step
807.” Id. at 6:61-62. Additionally, after “the location and rotation of the processed frame within
the display are determined, the frame may be cropped, discarding unneeded pixels, as in step 808.”
Id. at 6:65‐67. Next, the processed video frame “may be displayed in a heads up display, as in step
809.” Id. at 7:16‐17. As step 810 illustrates, “if another frame of video is set to be received (i.e.,
the display is still on), then the process repeats for each new frame, returning to step 802.” Id. at
7:30‐33.
It is evident that the language seamlessly transitions from video image to frames. In other
words, Figure 8 makes clear that the thing being generated, processed, and displayed is referenced
synonymously in the specification as being both images and video frames. For instance, in
explaining step 808, the specification states:
The frame may be resized in order to map the video information onto the pixels that
will ultimately be used in a heads up display. This step may be necessary if the
video images produced by a video source are larger than needed for display. For
example, if a video image initially has a field of view of 8 degrees horizontal and 6
degrees vertical, it may be cropped down to 4 degrees horizontal and 3 degrees
vertical, retaining the same center point. In this fashion, only a quarter of the image
33
is retained, but it constitutes the most relevant part of the image. Alternatively, the
video frame may need to be magnified or compressed in order to adjust for
differences in magnification between the visual field and the native video frame.
In addition, the frame may be enhanced by adding a border around the frame so as
to further distinguish it from the visual field for an observer.
Id. at 6:67-7:15.
In discussing resizing, the language of the specification demonstrates that the inventors
understood that images were equivalent to frames. The specification for the First Patent Family
begins with the general statement that “the frame” may need to be resized. Id. at 6:67-7:2. In the
very next sentence, the specification states that “video images” may need to be resized if the video
images are too large. Id. at 7:2-4. The specification then switches back to using the term “video
frames” which it states may need to be resized if too small. Id. at 7:9-12. The specification makes
a general statement that “the frame may be resized” then proceeds to discuss specifics of making
a “video image” smaller through cropping and making a “video frame” larger through magnifying.
Id. at 6:65-7:15. Plaintiff has not proffered any explanation for this interchangeable use. With no
apparent explanation to the contrary, the plain implication of this interchangeable use is that the
inventors view “video images” and “video frames” as the same object. See Edwards Lifesciences
LLC v. Cook Inc., 582 F.3d 1322, 1329 (Fed. Cir. 2009) (finding that the interchangeable use of
the words “graft” and “intraluminal graft” is “akin to a definition equating the two”).
The Second Patent Family, which shares a common inventor with the First Patent Family,
is no different and thus provides further support for Defendants’ construction. For instance, in
describing Figures 6A through 6F of the ’230 patent and explaining issues pertaining to parallax,
the ’230 Patent repeatedly uses the terms “video image” / “image” and “video frame” / “frame”
interchangeably. ’230 patent at 8:58-9:47. Figures 6A through 6F “illustrate positioning of one
video image within another based on data from inertial measurement units.” Id. at 2:50-53. For
34
example, Figure 6A shows the weapon and goggles pointing in the same direction, resulting in the
image from the weapon’s scope being placed directly in the middle of the image generated from
the goggles, which is showcased in Figure 6B. Id. at Figs. 6A-B, 8:66-9:2. In Figure 6C, the rifle
was yawed 8 degrees to the left of the goggles, which resulted in the scopes image being placed
negative 8 degrees from the centerline of the HUD display, as seen in Figure 6D. Id. at Figs. 6C-
D, 9:3-24. Similarly, in Figure 6E, the rifle has pitched upward 6 degrees, resulting in the rifle’s
image being shifted upward in relation to the center point generated by the goggles’ image, as
illustrated in Figure 6F. Id. at Fig. 6E-F, 9:25-30. In all the descriptions of these figures, the
Second Patent Family uses the words “image” or “view;” however, when the patent family
discusses parallax, the language seamlessly switches to the term “frame.” See, e.g., ’230 patent at
8:58-9:47. For instance, the ’230 patent states that when “processing a video frame from [the]
scope 17 [of the rifle], the location where the frame is placed may be slightly off, and a displayed
frame of video will not be aligned as perfectly as possible.” Id. at 9:36-39. (emphasis added).
Throughout this discussion, the ’230 patent uses the terms “frame” and “image” consistently to
reference to the thing being processed and displayed in the HUD. See, e.g., ’230 patent at 8:23-
26, 8:58-9:47.
Perhaps the best example demonstrating that the patents use “video images” / “image” and
“video frames” / “frames” interchangeably is contained in Figure 5A. Figure 5A through B
comprise a high-level flow chart which explains one possible operation by which video images are
registered. Id. at Fig. 5A, 7:46-8:57. In describing the step in Figure 5A where “video” is received
from “goggles and scope,” the specification states that, “[i]n block 107, a computer receives data
for a video frame (i.e., a scope image) from scope 17.” Id. at 8:23-26. In addition to
interchangeable use of words, patents may also define terms by implication by using the phrase
35
“i.e.” See Interval Licensing LLC v. AOL, Inc., 766 F.3d 1364, 1373 (Fed. Cir. 2014) (noting that
a phrase preceded by “i.e.” may indicate that the inventor’s intent to cast it as a definition);
Edwards Lifesciences LLC, 582 F.3d at 1334 (Fed. Cir. 2009) (finding the use of “i.e.” in the
specification “signals an intent to define the word to which it refers . . .”); but see Dealertrack,
Inc. v. Huber, 674 F.3d 1315, 1326 (Fed. Cir. 2012) (finding “i.e.” to not be definitional where
such a reading would exclude a preferred embodiment from the claim's scope or where a
“contextual analysis” of the patent indicates that “i.e.” is used in an exemplary rather than
definitional way).
Here, the term “i.e.” was used to equate “image” and “frame.” The use of i.e. in Figure 5A
is not merely incidental. The patents’ use of “video frames” / “frames” as interchangeable with
“video images” / “images” coupled with the use of “i.e.” clearly indicate that “video frame” and
“scope image” are intended to be used synonymously throughout the patents. See OpenTV, Inc. v.
Apple, Inc., 2015 WL 3544845, at *11 (N.D. Cal. June 5, 2015) (finding that specification’s
repeated and consistent usage of the term “redraw” as meaning “call low level graphics routine”
in conjunction with the specification’s use of “i.e.” clearly indicated that the patent intended to
define “redraw” as “call low level graphics routine”).
Notwithstanding, the evidence contained within the specifications of the First and Second
Patent Families, SAIC argues that the claimed “video images” are not equivalent to “video
frames.” Pl.’s Opening Cl. Constr. Br. at 10-14; Pl.’s Responsive Cl. Constr. Br. at 2-5; Pl.’s Supp.
Cl. Constr. Br. (ECF No. 149) at 12. First, SAIC argues that Defendants’ proposed construction
would destroy the claims’ distinction between partial frame display and complete frame display.
Pl.’s Opening Cl. Constr. Br. at 12-13. Second, SAIC argues that its proposed construction is
supported by the prosecution history. Pl.’s Opening Cl. Constr. Br. at 11-14.
36
Each of these contentions is without merit. First, SAIC’s concern that Defendants’
construction would destroy the claims’ distinction between partial frame display and complete
frame display is misplaced. By their plain language, the claims fall into two broad categories: (1)
claims that state part of the video image may be displayed; (2) claims that state the whole image
must be displayed. Notably, the claims that permit the display of part of the image say so
expressly. Compare ’230 patent, claim 15 (“A method comprising . . . (e) displaying at least a
portion of the first video source image and at least a portion of the second video source
image . . .”) (emphasis added) and ’752 patent, claim 7 (“A method comprising . . . displaying at
least a portion of the first image and the selected part of the second image . . .”) with ’012 patent,
claim 1 (“A method of registering video images with an underlying visual filed comprising the
steps of: . . . (3) displaying the video images . . . on the transparent display . . .”) and ’103 patent,
claim 1 (“A system comprising: . . . a computer adapted . . . to display the video images, on the
transparant [sic] display . . .”). Defendants’ proposed construction respects this distinction and
gives meaning to this claim language. Construing “video images” to mean “digital, analog, or
nonstandard video frames” permits partial frame display, and also allows for full frame display
when the claims require full frame display.
By contrast, SAIC’s proposed use of the term “electronic data” does not comport with the
claim language or the specifications. First, claim construction “must give meaning to all the words
in [the] claims.” Funai Elec. Co. v. Daewoo Elecs. Corp., 616 F.3d 1357, 1372 (Fed. Cir. 2010)
(internal quotation marks omitted) (quoting Exxon Chemical Patents, Inc. v. Lubrizol Corp., 64
F.3d 1553, 1557 (Fed. Cir. 1995)). Construing “video images” to encompass all types of electronic
data would destroy the distinction between partial and whole display by importing the ability to
represent only part of the frame into the definition of the term “video image.” If this were the case,
37
every claim would allow for partial frame display, regardless of its actual language. SAIC’s
construction would render the claims’ “at least a portion” language meaningless. Moreover, claim
1 of the ’012 and ’103 patents refer to the “boundaries of the displayed video images are in
registration with boundaries of portions of the visual field represented by the displayed video
images.” ’012 patent at 10:8-10; ’103 patent at 10:26-28. The notion of boundaries within the
claim supports Defendants’ proposed construction because frames are inherently bound; whereas,
electrical data is not. Therefore, Defendants’ construction is more consistent with the claim
language. See Trustees of Columbia Univ. v. Symantec, 811 F.3d 1359, 1366 (Fed. Cir. 2016)
(“[T]he construction that stays true to the claim language and most naturally aligns with the
patent’s description of the invention will be, in the end, the correct construction.” (internal
quotation and citation omitted)).
Second, SAIC’s construction would exclude “analog” images from the term video image
because nothing in the patent indicates that analog images are “electrical data.” For instance, the
patents state that video feeds may be delivered in numerous formats, including standard formats,
such as analog formats like NTSC or PAL, digital formats like MPEG, or any non-standard format.
’012 patent at 4:43-46; ’103 patent at 4:48-51; ’230 patent at 4:6-9; ’752 patent at 4:20-23. Frames
received in an analog format may require conversion into a digital format for processing. ’012
patent at 6:35-37; ’103 patent at 6:41-43. This reference to conversion into a digital format
indicates that the image being received may not always be “electronic.” As SAIC’s construction
would exclude analog uses because it fails to recognize the conversion of analog data to a digital
format, it would be improper to adopt its proposed construction. See Vitronics, 90 F.3d at 1583(“[A]n interpretation [which excludes a preferred embodiment] is rarely, if ever, correct . . . .”
(citations omitted)).
38
Finally, the prosecution history cited by SAIC does not support SAIC’s proposed
construction. During the prosecution of the ’230 patent, examiners originally rejected claim 1 as
unpatentable over the prior art, specifically Azuma’s LCD display. Pl.’s Opening Cl. Constr. Br.
Ex. E, ’230 Patent File History, May 30, 2012 Appeal Br. (ECF No. 90-8) at 4. SAIC appealed
the examiner’s rejection arguing that Azuma does not teach the disputed limitation in claim 1 of
the ’230 patent; rather, SAIC argued that the ’230 patent teaches that the “first video source [is]
configured to generate images representing portions of an external environment.” Id. at 7. To
support its appeal, SAIC noted, inter alia, that Azuma focused on optical images while claim 1 of
the ’230 patent is related to an image from a camera. Id. at 8-9. On appeal, the Patent Trial and
Appeal Board (PTAB) agreed with SAIC, reasoning that:
Although Azuma’s optical display may generate optical images, the disputed
limitation of a video source, when read in light of the Specification, must generate
video images, e.g., electrical signals or data. Therefore, it is insufficient that
Azuma’s display generates optical images; this fact, of itself, does not transform
Azuma’s display into a video source as required by claim 1. Furthermore, the
Examiner has erred by relying on Azuma’s single optical display to teach both the
first video source and video display of claim 1.
Pl.’s Opening Cl. Constr. Br. Ex. F, ’230 Patent File History, PTAB Decision on Appeal (ECF No.
90-9) at 5.
There is nothing cited in the ’230 patent prosecution history that supports SAIC’s assertion
that Defendants’ frame limitation is inappropriate. Instead, the cited prosecution history appears
to be more focused on distinguishing optical images from video images; it says nothing concerning
Defendant’s proposed “frame” construction.
SAIC’s reliance on the ’012 patent file history is also without merit. During the prosecution
of the ’012 patent, the patent examiner originally rejected SAIC’s claims based on a prior art night
vision goggle operational summary from STS Sensor Technology Systems, (STS). Pl.’s Opening
39
Cl. Constr. Br. Ex. G, ’012 Patent File History, July 24, 2009 (’012 Patent File History) (ECF No.
90-10) at 9-11. In response, SAIC discussed two combined images, provided below, generated by
the STS reference, which are not in registration:
Id. SAIC explained to the examiner that the above-depicted video images were not in registration
in part because “boundaries of the thermal [e.g., scope] image are substantially offset from (and
not in registration with) the boundaries of the visual field portion.” Id. at 11. These file history
remarks relate to the topic of registration, not video images, and nothing in these remarks addresses
the patent specification’s use of the terms “video image” and “video frame” interchangeably.
Accordingly, this Court construes “video images” as “digital, analog, or nonstandard video
frames.”
40
II. TRANSPARENT DISPLAY
SAIC Government Microsoft L3 11
a display that has the optical see-through N/A a display that allows
effect of display, or a display light from
being transparent or that allows some light the visual field to
translucent, to pass through it pass through it
allowing (i.e., is see-through)
simultaneous viewing when
of the underlying powered off
visual field and
other images or
information
Court’s Construction: “a display that has the effect of being transparent or translucent,
allowing simultaneous viewing of the underlying visual field and other images or information”
SAIC argues that in the context of the patents, the term “transparent display” should be
construed as “a display that has the effect of being transparent or translucent, allowing
simultaneous viewing of the underlying visual field and other images or information.” Pl.’s
Opening Cl. Constr. Br. at 34; Pl.’s Responsive Cl. Constr. Br. at 12-13; Pl.’s Supp. Cl. Constr.
Br. at 7; Pl.’s Responsive Cl. Constr. Br. to L3 at 1. SAIC argues that its construction is supported
by the specification which discloses “transparent display” in non-limiting terms and also includes
examples of both optical see-through and video see-through/functionally see-through. Pl.’s
Opening Cl. Constr. Br. at 36-37; Pl.’s Responsive Cl. Constr. Br. at 13-14; Pl.’s Supp. Cl. Constr.
Br. at 7-8; Pl.’s Responsive Cl. Constr. to L3 at 2. Thus, according to SAIC, a person of ordinary
11
L3 offers a slightly different construction for ’103 and ’012 patents. For the ’012 patent, L3
proposes construing “transparent display” as, “a display that allows light from the visual field to
pass through it.” L3 Opening Cl. Constr. Br. at 9. For the ’103 patent, L3 proposes construing
“transparent display” as, “a heads up display adapted for viewing of the visual field by a user of
the system wherein the HUD comprises a display that allows light from the visual field to pass
through it.” Id. The difference in construction seems to account for claim 1 of the ’012 patent
being a method claim, and, claim 1 of the ’103 patent being a system claim. However, L3’s
construction of “transparent display” is essentially a display that allows visual light to pass through
it. For ease of reference, this Court refers only to L3’s proposed ’012 patent claim construction.
41
skill in the art would understand the term transparent display to include video-generated displays
that have a transparent effect. Pl.’s Supp. Cl. Constr. Br. at 14-15.
The Government and L3 argue for a narrower interpretation of transparent display. The
Government interprets transparent display to mean an “optical see-through display, or a display
that allows some light to pass through it (i.e. is see-through) when powered-off.” Government’s
Opening Cl. Constr. Br. at 20, 27-28; Government’s Responsive Cl. Constr. Br. at 16, 20.
Similarly, L3 interprets transparent display to mean “a display that allows light from the visual
field to pass through it.” L3’s Opening Cl. Constr. Br. at 9; L3 Technologies, Inc.’s Responsive
Claim Construction Brief (ECF No. 152) (L3’s Responsive Cl. Constr. Br.) at 1. Though different,
L3’s construction agrees with the Government’s construction in principle. The gravamen of both
L3’s and the Government’s constructions is that transparent display is limited to an “optical see-
through display.” See Markman Hearing Tr. at 168:17-21 (L3), 188:22-190:22 (Government).
The parties’ dispute therefore boils down to whether “transparent display” is limited to so-
called “optical see-through displays” or whether the term encompasses so-called “video
see-through displays.” This Court agrees with SAIC’s construction that “transparent display”
encompasses “video see-through displays” because Defendants’ proposals would exclude
disclosed embodiments. See Vitronics, 90 F.3d at 1583-84. The term “transparent display”
appears in the First Patent Family claims only. The term appears in claims 1, 10, 12, and 17-18 of
the ’012 patent, and claims 1 and 9 of the ’103 patent. Nothing in the claim language limits
transparent display to optical see-through. Claim 1, for example, does not contain any express
limitation requiring the transparent display to operate only where light passes from the visual field
through the display. The dependent claims likewise do not include any such requirement.
42
With no evidence restricting transparent displays to optical see-through displays in the
claims themselves, both L3 and the Government rely heavily on Figure 6 of the ’012 patent to
support their constructions. L3’s Opening Cl. Constr. Br. at 10-11; Government’s Opening Cl.
Constr. Br. at 22-23; L3’s Responsive Cl. Constr. Br. at 5-7. Figure 6 is a block diagram depicting
the heads-up display (HUD) and a video assembly, both of which capture a view of the visual field.
’012 patent at Fig. 6. The HUD and video assembly are connected by sensors to a computer, and
the computer is then connected to a beam combiner. Id. Finally, the figure depicts an arrow
directly from the beam combiner to the viewer’s eye. Id. The parties agree that this particular
embodiment is optical see-through. See Government’s Opening Cl. Constr. Br. at 26; L3’s
Opening Cl. Constr. Br. at 5, 10-11; Markman Hearing Tr. 198:19-25 (SAIC). Defendants contend
that the invention is limited to this embodiment. They state that the description of Figure 6 limits
the term transparent display to an optical see-through display. Government’s Opening Cl. Constr.
Br. at 24-26; L3’s Responsive Cl. Constr. Br. at 5.
However, this Court cannot limit a transparent display to a single embodiment. Verizon
Servs. Corp. v. Vonage Holdings Corp., 503 F.3d 1295, 1305 (Fed. Cir. 2007). Moreover, “a claim
interpretation that excludes a preferred embodiment from the scope of the claim is rarely, if ever,
correct.” MBO Labs., Inc. v. Becton, Dickinson, & Co., 474 F.3d 1323, 1333 (Fed. Cir. 2007)
(internal quotation and citation omitted); EPOS Techs. Ltd. v. Pegasus Techs. Ltd., 766 F.3d 1338,
1347 (Fed. Cir. 2014) (holding that a claim construction is incorrect when it “reads out” preferred
embodiments).
Importantly, the ’012 patent expressly discloses two embodiments that use a video
see-through display. First, the ’012 patent expressly discloses an embodiment in which
the image produced in the visual field of a display is captured by a second video
camera. This second video feed or goggle image, along with the video images from
43
the video source, are both fed into a computer for initial digital processing. . . .
[T]he video image may be displayed in the heads up display alone, or the heads up
display may be filled with the resultant combination of video image and goggle
image.
’012 patent at 9:36-52.
At oral argument, L3 argued that this embodiment does not mean that the HUD can be
functionally see-through because the specification discloses that the HUD will be filled with a
single image. Markman Hearing Tr. at 179:4-181:19. This argument ignores the specification
statement that the single, resultant image is created by combining the video image and the goggle
image. ’012 patent at 9:49-51. There is no reason why such an embodiment would arbitrarily
exclude a video see-through display when the two-video system provides capability for a video
see-through display. Indeed, if this embodiment required that the video image for a goggle image
overlay an optically transparent display of that same goggle image, as L3 suggests, then such an
embodiment would require that three images be aligned instead of two—which is something that
the invention does not contemplate.
Second, the ’012 patent also states that,
[i]n other alternative embodiments, the heads up display need not be connected to
the viewer, as through a pair of night vision goggles. . . . The current state of the
art uses two screens, one for navigation and one for aiming the weapon. A robot
operator uses one screen to drive the robot and acquire targets, then refers to an
adjacent screen to aim and fire the weapon. Registering the weapon video image
to the navigation screen in a manner similar to an infantryman garners similar
advantages for the robot operator.
’012 patent at 5:23-25, 5:33-39. In such an embodiment, a robot is remote from its operator, who
uses a navigation screen to control the robot. Id. The Government’s and L3’s proposals would
exclude any variation of such an embodiment because light does not pass directly from the robot’s
visual field to the remote operator’s display.
44
Considering the specification’s disclosures, a POSITA (under either party’s proposed
standard) would understand that the inventions in the ’012 and ’103 patents use “transparent
display” as encompassing “video see-through” displays. Such an interpretation would also be
consistent with the inventions’ intended purpose of displaying video images in registration with
an underlying visual field. See ’012 patent at Abstract, 1:6-9; 2:32-38; claims 1, 17.
Notwithstanding the evidence in the specification, the Government and L3 also argue that
“during prosecution of the application issuing as the ’103 patent, for purposes of allowance, the
applicant authorized the following narrowing amendment to application claim 1, replacing (among
other things) ‘in the HUD’ with ‘on the transparent display[.]’” Government’s Opening Cl. Constr.
Br. at 24; L3’s Opening Cl. Constr. Br. at 9 (both quoting Examiner’s Amendment with Notice of
Allowability (ECF No. 63-3) at 127-29). The Government’s and L3’s construction is supported
by the ’103 patent’s prosecution history. Government’s Opening Cl. Constr. Br. at 24-25; L3’s
Responsive Cl. Constr. Br. at 8-10. Defendants argue that by making this amendment, SAIC ceded
from the claim scope a video see-through display. Government’s Opening Cl. Constr. Br. at 24-
25; see also L3’s Opening Cl. Constr. Br. at 9-10.
However, Defendants fail to articulate how the amendment results in a clear disavowal.
Nor could they. The law “requires that the alleged disavowing actions or statements made during
prosecution be both clear and unmistakable.” Aylus Networks, Inc. v. Apple Inc., 856 F.3d 1353,
1359 (Fed. Cir. 2017) (internal citations omitted). Here, no evidence exists on the record showing
why this amendment was made. While statements, including amendments, made during patent
prosecution can amount to prosecution disclaimer, the disavowal must not be ambiguous. See Avid
Tech., Inc. v. Harmonic, Inc., 812 F.3d 1040, 1045 (Fed. Cir. 2016) (“Where the alleged disavowal
is ambiguous, or even ‘amenable to multiple reasonable interpretations,’ [the Federal Circuit has]
45
declined to find prosecution disclaimer.” (quoting Cordis Corp. v. Medtronic AVE, Inc., 339 F.3d
1352, 1359 (Fed. Cir. 2003)) (citing Omega Engineering, Inc. v. Raytek Corp., 334 F.3d 1314,
1323-24 (Fed. Cir. 2003) (finding multiple reasonable interpretations of remarks made by the
inventor in patent prosecution so disavowal did not occur)). Replacing “in the HUD” with “on the
transparent display” alone says nothing about the scope of the term “transparent display.” Simply
put, the mere existence of an amendment does not demonstrate disavowal.
Indeed, a complete look at the prosecution history of the First Patent Family as a whole
indicates that there was no intended and clear disavowal of video see-through displays. Other
excerpts from the prosecution history of the ’012 patent indicate that SAIC neither excluded a
system that presented both the weapon sight image and the goggle image as video nor required
light from the goggles’ visual field to pass through the display. During prosecution, the examiner
rejected claim 1 of the ’012 patent application as obvious over the combination of (1) an
operational summary for a pair of night vision goggles made by Sensor Technology Systems
(STS) and (2) U.S. Patent No. 7,277,118 (Foote). Pl.’s Supp. Cl. Constr. Br., Ex. T, ’012 FH,
09.04.2009 Final Rejection (ECF No. 149-3) at 3-4. To support his rejection, the examiner alleged
that the combination of STS and Foote taught a multi-camera system providing video images. Id.
at 4. This rejection indicates that the examiner viewed claim 1 of the ’012 patent as encompassing
multiple cameras. SAIC overcame this rejection, not by arguing that claim 1 employed only a
single camera, but by arguing, inter alia, that “Foote’s predication upon fixed cameras is contrary
to (and fundamentally incompatible with) the recitation in claim 1 of the video source and the
transparent display being independently movable about multiple axes.” Pl.’s Supp. Cl. Constr. Br.
Ex. U, ’012 FH, 12.03.2009 Arguments (ECF No. 149-4) at 8-9; see also Pl.’s Supp. Cl. Constr.
Br. Ex. V, 12.17.2009 Pre-Appeal Brief Request for Review (ECF No. 149-5) at 3. SAIC’s
46
argument thus differentiated claim 1 from the examiner’s asserted combination of STS and Foote
because the transparent display was independently movable from the video source but did not
disclaim using video system as the transparent display. Accordingly, the Court finds that the
prosecution history does not evidence a clear disavowal of a video see-through transparent display.
Defendant’s reliance on extrinsic evidence is also unavailing. Because the intrinsic record
clearly contemplates “transparent displays” beyond optical see-through displays, this Court need
not reference extrinsic evidence. See Summit 6, LLC v. Samsung Elecs. Co., Ltd., 802 F.3d 1283,
1290 (Fed. Cir. 2015) (“Although courts are permitted to consider extrinsic evidence, like expert
testimony, dictionaries, and treatises, such evidence is generally of less significance than the
intrinsic record.” (citing Phillips, 415 F.3d at 1317)). Furthermore, it is imperative that this Court
put far less weight, if any, on extrinsic evidence to the extent it contradicts the intrinsic record.
See Immunex Corp. v. Sanofi-Aventis U.S. LLC, 977 F.3d 1212, 1221-22 (Fed. Cir. 2020) (citing
Helmsderfer v. Bobrick Washroom Equip., Inc., 527 F.3d 1379, 1382 (Fed. Cir. 2008)) (holding
that it would be incorrect “to allow the extrinsic evidence . . . to trump the persuasive intrinsic
evidence”).
Here, the Government and L3 cite to several dictionary definitions, testimony from Dr.
Ronald Azuma, and prior art to support their constructions that would limit the transparent display
to optical see-through displays. See Government’s Opening Cl. Constr. Br. at 28; L3’s Opening
Cl. Constr. Br. at 13, 24; L3’s Responsive Cl. Constr. Br. at 10-12. With respect to dictionaries,
the Defendants rely on the following definitions:
(1) The Government uses Merriam-Webster’s Collegiate Dictionary (11th ed.),
which defines “transparent,” in relevant parts, as “1a(1): having the property of
transmitting light without appreciable scattering so that bodies lying beyond are
47
seen clearly” or “1b: fine or sheer enough to be seen through,” “b: easily detected
or seen through.” Government’s Opening Cl. Constr. Br. at 28 (citing SAIC’s
Motion to Substitute Exhibits Ex. A, Merriam-Webster’s Collegiate Dictionary
(11th ed. 2004) at 1330 (ECF No. 69-1) at 3).
(2) L3 uses Merriam Webster’s Collegiate Dictionary (10th ed.), which defines
“transparent” as “having the property of transmitting light without appreciable
scattering so that bodies lying beyond are seen clearly.” L3’s Opening Cl. Constr.
Br. Ex. 1 (Merriam-Webster’s Collegiate Dictionary (10th ed. 2000) at 1251)
(ECF No. 148-1) at 5.
(3) L3 also uses Wiley Electrical and Electronics Dictionary, which defines
“transparent,” in relevant part, as “1. A body, material or medium which freely
passes radiant energy, such as light, or sound.” L3’s Opening Cl. Constr. Br. Ex.
6 (Wiley Electrical and Electronics Dictionary (2004) at 801-02) (ECF No. 148-
1) at 26.
The Government and L3 also rely on the testimony of Dr. Azuma and on prior art to
distinguish between optical see-through and video see-through heads up displays. See
Government’s Opening Cl. Constr. Br. at 26-27 (citing Joint Claim Construction Statement Exhibit
12, Chung, J.C., et al., “Exploring Virtual Worlds with Head-Mounted Displays,” appeared in
Non-Holographic True 3-Dimensional Display Technologies, SPIE Proceedings, Vol. 1083, Los
Angeles, CA, January 15-20, 1989 (ECF No. 64-12) at 5) (differentiating between “optical see-
though HMD” and “opaque head-mounted display”); Pl.’s Responsive Cl. Constr. Br. Ex. S,
Ronald Azuma Deposition Transcript (ECF No. 96-3) (Azuma Dep.) at 160:21-161:2 (“Q.
. . . Would you consider a video see-through as it's used here in your dissertation -- excuse me.
48
Video see-through HMD, would you consider that to be a transparent display? A. I think we had
that discussion already. Generally speaking, no, because if the power was cut, then the user would
effectively be blinded.”).
The dictionary definitions, Dr. Azuma’s testimony, and the prior art referenced by
Defendants are contrary to the intrinsic record and therefore fail to support Defendants’ proposed
construction. Even if these dictionaries, expert testimony, and prior art did establish that
video-see-through displays are generally not considered “transparent,” it would be of no moment;
under Phillips, the plain meaning is not the meaning of the term in the abstract but is rather the
plain meaning as understood by a POSITA after reading the patents. 415 F.3d at 1313
(“Importantly, the person of ordinary skill in the art is deemed to read the claim term not only in
the context of the particular claim in which the disputed term appears, but in the context of the
entire patent, including the specification.”). When a term as used in patent conflicts with the
ordinary or conventional usage of that term, the patent’s use of the term must prevail. For instance,
in Honeywell Int'l, Inc. v. Universal Avionics Sys. Corp., 493 F.3d 1358, 1361 (Fed. Cir. 2007),
the Federal Circuit was tasked with interpreting the term “heading” in the context of aircraft
navigation. There, the parties agreed that, in the context of aircraft navigation, ‘“[h]eading’
ordinarily refers to the direction in which an object is pointing.” Id. However, the Federal Circuit
found that, based on a patent figure, “the patentees used the term ‘heading’ in a manner different
from its ordinary meaning.” Id. Relying on Phillips, the Federal Circuit held that “[w]hen a
patentee defines a claim term, the patentee's definition governs, even if it is contrary to the
conventional meaning of the term.” Id. (citing Phillips, 415 F.3d at 1321). Here, the patents
disclose embodiments that include both optical see-through and video see-through. See ’012
patent at 5:44-6:2, 7:16-19; ’103 patent at 5:49-6:7, 7:22-25. To the extent the cited dictionary
49
definitions, prior art references, and expert testimony contradict these disclosed embodiments, this
Court cannot rely on such definitions, references, and testimony, and declines to do so. See
Immunex Corp. v. Sanofi-Aventis U.S. LLC, 977 F.3d 1212, 1221-22 (Fed. Cir. 2020) (citing
Helmsderfer v. Bobrick Washroom Equip., Inc., 527 F.3d 1379, 1382 (Fed. Cir. 2008)) (holding
that it would be incorrect “to allow the extrinsic evidence . . . to trump the persuasive intrinsic
evidence”).
For these reasons, this Court construes “transparent display” as “a display that has the effect
of being transparent or translucent, allowing simultaneous viewing of the underlying visual field
and other images or information.”
III. OVERLAY
SAIC Government Microsoft L3
Plain and ordinary N/A N/A to place on top or in
meaning: front of
e.g., Overlaying (e.g., superimpose)
(plain and
ordinary meaning):
positioned over or
upon.
Overlay/overlays
(plain and
ordinary meaning):
are over
or upon/is over or
upon
Court’s Construction: Plain and ordinary meaning, e.g., Overlaying: positioned over or upon. /
Overlays/overlay: are over or upon/is over or upon.
SAIC’s and L3’s proposed constructions of the terms “overlay,” “overlays,” and
“overlaying” are remarkably similar, but are not identical. SAIC contends that the “overlay” terms
are unambiguous because such terms are used in accordance with their plain and ordinary meaning
50
in the patents—“positioned over or upon” or “are over or upon/is over or upon.” Pl.’s Supp. Cl.
Constr. Br. at 16. SAIC further argues that L3’s proposed construction “to place on top or in front
of (e.g., superimpose)” as a synonym for “overlay” adds nothing of consequence to the term’s
meaning and would erroneously read limitations into the patent claims. Id. at 18 (citing
ActiveVideo Networks, Inc. v. Verizon Comm’ns, Inc., 694 F.3d 1312, 1324-25 (Fed. Cir. 2012));
Pl.’s Responsive Cl. Constr. Br. to L3 at 7. Essentially, SAIC urges the Court to adopt its proposed
plain and ordinary meaning or construction of the term “overlay” because it “encompasses the full
scope of the disclosed embodiments consistent with the inventions’ purpose;” whereas, L3’s
proposed construction would inappropriately limit the scope of the patent claims at issue. Pl.’s
Supp. Cl. Constr. Br. at18; Pl.’s Responsive Cl. Constr. Br. to L3 at 7-8.
L3 argues that “overlay” should be construed to mean “on top or in front of (e.g.,
superimposed).” L3’s Responsive Cl. Constr. Br. at 14. L3’s proposed construction imports a
directional component into the definition of “overlay” that SAIC’s construction does not require.
See L3’s Opening Cl. Constr. Br. at 5-8. L3 asserts this definition comports with how “overlaying”
is described in the patents. L3’s Responsive Cl. Constr. Br. at 14-15.
The parties appear to agree—and the Court concurs—the concept of “superimposing” is
contained within the parties’ respective claim constructions, which L3 and SAIC each contend
encompass the “plain meaning” of the “overlay” terms. See, e.g., L3’s Opening Cl. Constr. Br. at
8-9; L3’s Responsive Cl. Constr. Br. at 14-16; Pl.’s Cl. Constr. Presentation at 88 (“SAIC offered
to construe this term as ‘plain and ordinary meaning (e.g., superimpose)’”); L3’s Cl. Constr.
Presentation at 33 (noting SAIC’s proposed stipulation encompasses “superimpose”). The parties’
dispute appears instead to center around the importation of directional requirements into the
“overlay.”
51
The Court agrees with SAIC that overlay is not limited to placing one image “on top of”
or “in front of” another. The overlay terms appear in claims 1 and 17 of the ’012 patent; claim 1
of the ’103 patent; and claims 1, 15, and 29 of the ’230 patent. Claim 1 of the ’012 patent recites
a method of displaying video images overlaying portions of a visual field. ’012 patent at 9:63-
10:10. The ’012 patent states:
1. A method of registering video images with an underlying visual field comprising
the steps of:
(1) determining a source orientation of a video source providing a video feed
containing data for a series of video images representing portions of a visual
field;
(2) determining a display orientation of a transparent display overlaying the
visual field, wherein the video source and the transparent display are
independently movable about multiple axes; and
(3) displaying the video images in positions on the transparent display that
overlay portions of the visual field represented by the displayed video images,
wherein boundaries of the displayed video images are in registration with
boundaries of portions of the visual field represented by the displayed video
images.
Id.In claim 1 of the ’012 patent, “overlaying” is used to distinguish the transparent display
from the “underlying visual field.” 12 As used in claim 1, the terms “overlaying” and “underlying”
describe the positions of the transparent display and the visual field relative to one another.
However, nothing in claim 1 limits “overlaying” through an objective perspective outside both the
transparent display and the visual field such that the overlaying image must always be viewed as
“on top of” or “in front of” another image.
In the Second Patent Family, only the ’230 patent uses the term overlay. 13 The ’230 patent
12
Claim 1 of the ’103 patent uses the term “overlay” consistent with the rest of the First Patent
Family. See ’103 patent at 10:20-29.
13
The ’752 patent uses the term “replace” rather than “overlay.” However, at oral argument on
the Government’s motion to dismiss, SAIC’s counsel noted that the terms serve a similar function.
52
uses “overlay” to describe how two video source images are displayed, rather than to distinguish
between the video feed and visual field as in the First Patent Family. Claim 1 of the ’230 patent
is illustrative and recites in relevant part:
(e) display at least a portion of the first video source image and at least a portion
of the second video source image such that the second video source image
portion overlays a corresponding region of the first video source image portion,
wherein the corresponding region represents a portion of the external
environment represented in the second video source portion.
’230 patent at 24:45-51.
Like the claims in the ’012, ’103, and ’230 patents, the First Patent Family’s specification,
which is incorporated by reference in ’230 patent’s specification, occasionally describes an image
as “in front of the visual field” when describing certain embodiments. ’012 patent at 5:57-67;
7:21-24. However, nothing in the specifications indicate that “overlay” is necessarily limited to
placing an image in front of or on top of another. Therefore, the importation of such limitations
from the specification is impermissible. See Phillips, 415 F.3d at 1323.
Moreover, L3’s construction would impermissibly exclude other embodiments. See MBO
Labs., 474 F.3d at 1333 (“A claim interpretation that excludes a preferred embodiment from the
scope of the claim is rarely, if ever, correct.” (internal quotation marks and citation omitted)). For
instance, the ’012 patent specification references alternative embodiments where “the heads up
display may be filled with the resultant combination of the video image and the goggle image.”
’012 patent at 9:37-51. L3’s proposed limitation would exclude this embodiment because placing
one image in front of another requires that there be two distinct images that, when viewed from a
specific (though unidentified) perspective, are directionally related such that one is “on top or in
Jan. 3, 2018 Oral Argument Tr. (ECF No. 18) at 68:4-17 (“[The ’752 patent] talks about replacing
an image with a portion of another – replace a portion of an image with a portion of another image.
That’s different. That’s different language. It’s a different process. It may get you to the same
end, but you asked about the how, the how matters. Those are different hows, Your Honor.”).
53
front of” the other. L3’s Responsive Cl. Constr. Br. at 16. Specifically, L3’s limiting definition
would exclude such an embodiment from the scope of the patent claim because the HUD is “filled
with” a single resultant image rather than one image positioned “on top or in front of” another.
See ’012 patent at 9:49-51.
Nor is L3’s reliance on dictionary definitions persuasive. As noted, because the intrinsic
record is clear that “overlay” is not limited to “in front of’ or “on top of,” this Court need not
reference extrinsic evidence. See Phillips, 415 F.3d at 1317-18; Summit 6, LLC, 802 F.3d at 1290.
However, to the extent extrinsic evidence is at all relevant, it supports a broader construction than
L3 proposes.
To support its proposed construction, L3 relies on the following dictionary definitions of
the term “overlay.” L3’s Opening Cl. Constr. Br. at 8-9.
(1) Merriam Webster’s Collegiate Dictionary defines “overlay” as “to lay or spread
over or across.” L3’s Opening Cl. Constr. Br. Ex. 1 (Merriam-Webster’s
Collegiate Dictionary (10th ed. 2000) at 827) at 4.
(2) New Penguin Dictionary of Computing (2001) defines “overlay” in the context
of graphics as “superimpos[ing] one image over another.” L3’s Opening Cl.
Constr. Br. Ex. 2 (New Penguin Dictionary of Computing (2001) at 352) (ECF
No. 148-1) at 9.
(3) The New Oxford American Dictionary defines “overlay” as “lie on top of: a third
screen which will overlay the others.” L3’s Opening Cl. Constr. Br. Ex. 3 (The
New Oxford American Dictionary (2nd ed. 2005) at 1213) (ECF No. 148-1) at
13.
54
(4) Newton’s Telecom Dictionary defines “overlay” as “The ability to superimpose
computer graphics over a live or recorded video signal and store the resulting
video image on videotape. It is often used to add titles to videotape.” L3’s
Opening Cl. Constr. Br. Ex. 5 (Newton’s Telecom Dictionary (16th ed. 2000) at
622) (ECF No. 148-1) at 21.
In Phillips, the Federal Circuit explained that dictionaries may be useful in determining the
meaning of terms. 415 F.3d at 1318. Despite this allowable use of dictionary definitions, the
Federal Circuit has warned that courts should not rely on dictionary definitions that are “entirely
divorced from the context of the written description.” Id. at 1321. The Federal Circuit thus
concluded that:
[J]udges are free to consult dictionaries and technical treatises “at any time in order
to better understand the underlying technology and may also rely on dictionary
definitions when construing claim terms, so long as the dictionary definition does
not contradict any definition found in or ascertained by a reading of the patent
documents.”
Id. at 1322-23 (quoting Vitronics, 90 F.3d at 1585 n.6). Here, the New Oxford American
Dictionary, which appears to be a general use dictionary, defines “overlay” as meaning to “lie on
top of: a third screen which will overlay the others.” L3’s Opening Cl. Constr. Br. Ex. 3 at 13. To
be sure, this definition supports L3’s construction (and is encompassed by SAIC’s construction)
and is consistent with some embodiments contained within the patents. ’012 patent at 5:57-67;
7:21-24. However, the other dictionaries to which L3 cite do not support limiting the term
“overlay” as L3 suggests. Indeed, Newton’s Telecom Dictionary and New Penguin Dictionary of
Computing, which are technical in nature, appear to undermine L3’s proposed claim limitations.
Newton’s Telecom Dictionary defines “overlay” as “[t]he ability to superimpose computer
graphics over a live or recorded video signal and store the resulting video image on videotape.”
55
L3 Opening Cl. Constr. Br. Ex. 5 at 21. Similarly, New Penguin Dictionary of Computing defines
“overlay” in the context of graphics as “superimpos[ing] one image over another.” L3’s Opening
Cl. Constr. Br. Ex. 2 at 9. These definitions, consistent with the intrinsic evidence, do not require
that an image is placed only “in front of” or “on top of” another. Newton’s Telecom Dictionary’s
reference to a “resulting video image” is also consistent with the ’012 patent specification’s
reference to a preferred embodiment which combines images from two video feeds into one
resultant image using digital processing. See ’012 patent at 9:37-51. Accordingly, SAIC’s
proposed construction is consistent with both the intrinsic and extrinsic evidence cited by L3.
For these reasons, this Court construes the “overlay” terms in accordance with their plain
meaning as, “e.g., Overlaying: positioned over or upon.” / “ Overlay/overlays: are over or upon/is
over or upon.”
IV. “BASED ON A COMPARISON OF DATA FROM THE FIRST AND SECOND
VIDEO SOURCE IMAGES”
SAIC Government Microsoft L3
Plain and ordinary N/A N/A based on [a] the
meaning comparison of
image data (e.g.,
content and
contrast) from the
first and
second video source
images
Court’s Construction: “based on a comparison of image data (e.g., content and contrast) from
the first and second video source images”
In its most recent filing, L3 offers the construction “based on [a] the comparison of image
data (e.g., content and contrast) from the first and second video source images” for the “based on
a comparison of data from the first and second video source images” term. L3’s Responsive Cl.
56
Constr. Br. at 17 (brackets and strikethrough in original). 14 L3 argues that the context of the claims
and specification demonstrate that the “data” being compared is “image data (e.g., content and
contrast)” and not “motion data” or “orientation data.” L3’s Opening Cl. Constr. Br. at 14-17. In
other words, L3’s construction seeks to clarify “the actual content of the images – [sic] whether it
be their greyscale, contrast, PSR, or some similar measure—is what must be compared, not data
unrelated to the content of the images, such as the separately claimed motion/orientation data.”
L3’s Responsive Cl. Constr. Br. at 18.
SAIC argues that L3’s proposed construction would unduly restrict the claims to less than
their full scope. Pl.’s Responsive Cl. Constr. Br. to L3. at 8. Specifically, SAIC contends that,
[w]hile content and contrast are two types of data used when comparing data from
a first and second sources of images as part of image registration, the express claim
language—i.e., data from the first and second video source images—is
straightforward and L3’s proposal to rewrite it should be rejected.
Pl.’s Supp. Cl. Constr. Br. at 21-22 (emphasis in original).
This Court agrees with L3 that the “data” in the phrase “based on a comparison of data”
refers to “image data (e.g., content and contrast).” The language of the claims read in context of
the entire patent are of primary importance. See Phillips, 415 F.3d at 1312 (citing Merrill v.
Yeomans, 94 U.S. 568, 570 (1876)). Here, the language of the claims read in light of the
specification are clear. The ’230 patent claims a process of registering images from two different
video sources through two different comparisons—the first being based on a comparison of motion
14
Initially, L3 proposed to construe “based on a comparison of data from the first and second video
source images” as “based on the comparison of image data (e.g., content and contrast) from the
first and second video source images.” L3’s Opening Cl. Construction Br. at 14. However,
Plaintiff took issue with L3’s initial construction language of “the comparison” as opposed to “a
comparison” because it would allegedly introduce “an otherwise non-existent antecedent basis
problem” to the claim language. Pl.’s Responsive Cl. Constr. Br. to L3 at 11. L3’s revised, current
construction uses “a” instead of “the” to avoid an antecedent basis issue. L3’s Responsive Cl.
Constr. Br. at 18.
57
data from the video source and the second being based on a comparison of the content of the
images themselves. See, e.g., ‘230 patent at Abstract.
The phrase “based on a comparison of data from the first and second video source images”
is contained in step (d) of claims 1, 15, and 29 term of the ’230 patent. The distinction between
motion data and image data is illustrated by claim 1 of the ’230 patent. In claim 1, the patentee
outlines the invention as “[a] system, comprising: a first video source configured to generate
images representing portions of an external environment [and] a second video source, movable
independent of the first video source configured to generate images representing portions of the
external environment . . . .” ’230 patent at 24:25-30. These video sources are connected to a
controller and video display “wherein the controller is configured to (a) receive video images from
the first video source and from the second video source” then to “(b) receive motion data indicative
of motion of the first and second video sources.” Id. at 24:33-37. Once the controller receives
motion and image data from the first and second video sources, it then “(c) identif[ies], based on
the received motion data, a part of a first video source image that potentially represents a portion
of the external environment represented in a part of a second video source image . . . .” Id. at
24:38-41. Crucially, the system then “(d) evaluate[s], based on a comparison of data from the first
and second video source images, the identification performed in operation (c).” Id. at 24:42-44.
Finally, the system,
display[s] at least a portion of the first video source image and at least a portion of
the second video source image such that the second video source image portion
overlays a corresponding region of the first video source image portion, wherein
the corresponding region represents a portion of the external environment
represented in the second video source portion.
Id. at 24:45-51.
The structure of claim 1 and step (d)’s reference to “data from the first and second video
source image” demonstrates that the data used to perform this evaluation is the data received from
58
“video images” in step (a). Conversely, the data used to perform the evaluation in step (d) is not
the “motion data” described in step (b), but rather is the data from the images received in step (a).
See Becton, Dickinson & Co. v. Tyco Healthcare Grp., LP, 616 F.3d 1249, 1254 (Fed. Cir. 2010)
(“Where a claim lists elements separately, the clear implication of the claim language is that those
elements are distinct components of the patented invention.” (internal quotations and citations
omitted)).
Reading claim 1 in light of the invention’s intended purpose along with the specification
provides further support for L3’s construction. The First Patent Family claims a system in which
the images from two different sources are aligned using only orientation data. See, e.g., ’012
patent, claims 1, 17; ’103 patent, claim 1. By the time the ’230 patent was filed three years later,
however, the named inventors had realized that use of orientation data alone may pose problems.
’230 patent at 1:35-43 (identifying disadvantages of a system using only sensor data to match
images). Thus, the later-filed ’230 patent describes an improved two-step alignment method. This
method first uses data from motion sensors to help align images from two different sources and
then performs a second step of comparing the content of the images themselves; then, using that
comparison to evaluate whether the alignment is correct, adjusts the alignment as necessary. See,
e.g., id. at Abstract (“The sensor-based location is checked (and possibly adjusted) based on a
comparison of the images.”); 2:57-62 (“FIGS. 8A through 8K illustrate checking and/or correcting
an IMU-based position for one video image within another video image. FIGS. 9A through 9C
illustrate correction of an IMU-based position calculation based on image comparison results.”);
3:1-5 (“FIGS. 13A and 13B show a goggles images and scope image, respectively, used to describe
an alternative image comparison algorithm. FIG. 14 is a flow chart for an alternative image
comparison algorithm.”); 7:19-22 (“As discussed below, the location and rotation of weapon view
59
74 within user display 70 is determined by computer 30 based on output from sensors 13 and 18
and based on comparison of the scope image with the goggles image.”); 9:45-47 (“For larger
distances, image comparison position calculations (described below) compensate for errors caused
by parallax.”); 10:5-16 (“In block 117, the IMU-based calculation for position and rotation of
weapon view 74 within display 70 is checked using an image-based method.”). The ’230 patent
summary provides an overview of this two-step alignment process:
Data from the two images are then compared in order to evaluate the location
determined from the sensor data. The sensor-based location is either confirmed, or
a new location is found based on additional image comparisons. Once a location is
selected (either a confirmed sensor-based location or a location found using image
comparison), the two images are displayed such that the second source image (or a
portion of that image) overlays a corresponding portion of the first source image.
Locations obtained using image comparisons are used to calibrate (adjust) the
manner in which subsequent sensor-based locations are determined.
’230 patent at 2:6-17.
L3’s proposed construction gives proper meaning to the ’230 patent’s alignment process
by delineating the use of motion data for the first step of the alignment process and the use of
image data for the second step.
In contrast, SAIC cannot cite to any intrinsic or extrinsic evidence to contradict L3’s
construction; instead, SAIC argues in conclusory fashion that the phrase “based on a comparison
of data from the first and second video source images” needs no construction. Pl.’s Supp. Cl.
Constr. Br. at 21-22. Indeed, the portions of the specification SAIC references appear to support
L3’s position that step (d) “evaluates” the accuracy of the initial alignment derived from motion
data using the “content and contrast” of the images from the first and second video source. See
’230 patent at 10:21-26 (“To address these concerns, the relative orientation of goggles 11 and
scope 17 can be independently deduced by processing image data from image generator 57 and
60
scope 17 if there is sufficient image content and contrast and if similar imaging technologies (e.g.
microbolometers, CCD, etc.) are used.”).
Accordingly, the Court construes “based on a comparison of data from the first and second
video source images” to mean “based on the comparison of image data (e.g., content and contrast)
from the first and second video source images.”
V. “MOTION DATA”
SAIC Government Microsoft L3
“data indicative of Indefinite N/A Indefinite
motion, including at
least orientation data”
Court’s Construction: “data indicative of motion, including at least orientation data”
The Government contends that the term “motion data” is a word or term with no established
meaning outside the patents. Government’s Opening Cl. Constr. Br. at 15. Accordingly, the
Government argues that, because there is no definition or even mention of “motion data” in the
specification, a POSITA reviewing the specification and prosecution history at the time of the
invention would not be informed “with reasonable certainty” of the claimed invention’s objective
scope. Id. at 18. Moreover, according to the Government, SAIC’s construction of “motion data”
as “data indicative of motion, including at least orientation data” is circular because it includes
“data” and “motion” in the definition of “motion data,” then unjustifiably adds “including at least
orientation data,” opening the door for additional, undefined “data” to be swept up within the
scope. Id. at 17-19. The Government also contends that SAIC’s proposed construction would be
repetitive and nonsensical because replacing “motion data” with SAIC’s proposed construction (as
shown in underline) in claim 1, clause (b) of the ’230 patent would read as follows: “(b) receive
61
data indicative of motion, including at least orientation data indicative of motion of the first and
second video sources.” Id. at 16. The Government also relies on extrinsic evidence, particularly
the testimony of Ulrich Neumann, Ph.D., 15 Ronald Azuma, Ph.D., 16 and Gregory Welch, Ph.D.,17
to support its assertion. See id. at 15-19.
SAIC argues that motion data is readily understandable from the context of the patents and
the surrounding claim language. Pl.’s Responsive Cl. Constr. Br. at 9. Additionally, SAIC argues
that its construction is consistent with extrinsic evidence. Specifically, SAIC points to Dr. Welch’s
testimony that the patents teach utilization of orientation data to determine the location of the
weapon image within the goggles image. Pl.’s Responsive Cl. Constr. Br. at 10-11; see also Pl.’s
Opening Cl. Constr. Br. at 33-34.
The Court agrees with SAIC’s construction and holds that “motion data,” as used in the
patents, is not indefinite. The intrinsic evidence supports SAIC’s construction. This term only
appears in the Second Patent Family claims, particularly claims 1-3, 5, 15, 17, 19, 29, 31, and 33
of the ’230 patent, and claims 1-2, 7-8, and 13-14 of the ’752 patent. Importantly, in the ’230 and
’752 patents, this term appears for the first time in the claims and cannot be found anywhere in the
specification. The Government contends that “motion data” does not have a defined meaning
outside the patents, and thus the specification must define the term to provide reasonable certainty.
Government’s Opening Cl. Constr. Br. at 15 (quoting citing Acacia Media Techs. Corp. v. New
15
Dr. Neumann is the Government’s claim construction expert. Defendant’s Disclosure of Claim
Construction Expert (ECF No. 48).
16
Dr. Azuma is a non-party, fact witness subpoenaed by the Government; however, the parties
agree that he is also a “recognized pioneer and innovator in augmented reality.” Azuma Dep. at
100:22-103:2, 136:5-8.
17
Dr. Welch is SAIC’s claim construction expert. Disclosure of Claim Construction Expert (ECF
No. 49).
62
Destiny Internet Grp., 405 F. Supp. 2d 1127, 1136 (N.D. Cal. 2005)). The Government’s reliance
on Acacia is misplaced. Acacia involved a patent for a data transmission system in which the
plaintiff sought to define a “sequence encoder” as a “time encoder” to avoid an indefiniteness
ruling. 405 F. Supp. 2d at 1134-36. While “time encoder” was used in one embodiment, the
district court found no evidence that this was to be the only embodiment. Id. at 1136. Acacia held
that,
[i]f a patentee uses a broad undefined term (such as ‘[motion data]’) in claiming an
invention, when the validity of the patent is called into question in a legal
proceeding, the owner of the patent cannot avoid invalidity by adopting a more
limited definition (such as ‘[orientation data]’), unless that limitation can be fairly
inferred from the specification.
Id. Acacia is simply inapposite to the current case. SAIC is not seeking to limit “motion data” to
“orientation data” but rather is arguing that orientation data is one type of data that must always
be included within “motion data,” as that term is used within the patents. Pl.’s Cl. Constr. Br. at
32-33. More importantly, unlike the plaintiff in Acacia that sought to use an embodiment to limit
the scope of a claim, here, SAIC’s construction can fairly be inferred from the specification.
The Acacia court acknowledged that not all undefined terms are indefinite. 405 F. Supp.
2d at 1136. As illustrated by the Federal Circuit’s decisions in Bancorp Servs., L.L.C. v. Hartford
Life Ins. Co., 359 F.3d 1367 (Fed. Cir. 2004) and Network Commerce, Inc. v. Microsoft Corp., 422
F.3d 1353 (Fed. Cir. 2005), an undefined term with no specialized meaning in the field of the
invention is not indefinite where the meaning of the term is fairly inferred from the patent. In
Bancorp, the Federal Circuit addressed whether the phrase “surrender value protected investment
credits” in a patent that described a system for tracking the value of life insurance policies was
indefinite. 359 F.3d at 1372. The phrase was not defined in the patent and did not have an
established definition in the industry. Id. at 1372-73. However, the Federal Circuit held the term
was definite because (1) the phrase’s component terms had “well-recognized meanings which
63
allow the reader to infer the meaning of the entire phrase with reasonable confidence,” and (2) the
meaning of the phrase was “fairly inferable” from the specification and the dependent claims. Id.
at 1372-74. Likewise, in Network Commerce, the Federal Circuit held the term “download
component” definite, despite being undefined, where the patents provided sufficient context as to
how “download component” functioned in the claimed method. 422 F.3d at 1360-61.
Here, ’230 patent claim 1 states that the system is comprised of “a controller coupled to
the first and second video sources and to the display, wherein the controller is configured to . . .
(b) receive motion data indicative of motion of the first and second video sources . . . .” ’230
patent at 24:32-37. Step (b) of ’230 claim 1 identifies motion data. ’230 patent at 24:36-37. All
subsequent uses of “motion data” are derived from how motion data is used in step (b). See, e.g.,
Id. at 24:38-41.
Step (c) of claim 1 continues that the controller will “identify, based on the received motion
data, a part of a first video source image that potentially represents a portion of the external
environment represented in a part of a second video source image.” Id. at 24:38-41. Thus, step
(c) of ’230 claim 1 identifies how motion data is used.
Read in light of the specification, it is clear that, at minimum, the motion data in steps (b)
and (c) of claim 1 must measure or account for the relative orientation or alignment of both video
sources. As discussed supra Section IV, the Second Patent Family claims a system in which the
images from two different sources are aligned using only orientation data. See, e.g., ’012 patent,
claims 1, 17; ’103 patent, claim 1. The ’230 patent purports to correct errors associated with the
First Patent Family by adding an additional “evaluation” step to “check” the alignment initially
performed using orientation data. See, e.g., ‘’230 patent at 24:42-44. Thus, claim 1 steps (b) and
(c) are intended to incorporate the First Patent Family’s use of orientation data. For instance, in
64
describing the First Patent Family, the Second Patent Family discloses that “[s]ensors coupled to
the rifle and to the goggles provide data indicating movement of the goggles and rifle. [T]he sensor
data is then used to determine the relative orientation of the two sources and calculate a location
for the rifle image within the image seen through the goggles.” ’230 patent at 1:25-34. In
summarizing the invention, the ’230 patent uses almost identical language to describe the first step
of Second Patent Family’s two-step alignment process. The summary of invention states that “[i]n
at least some embodiments . . . , [s]ensors coupled to the two video sources provide data to the
computer that indicates the spatial orientations of those sources. Using the sensor data, the
computer determines a location for placing a video image (or a portion thereof) from a second of
the sources (e.g., a rifle-mounted source) in the video image from a first of the sources (e.g., a
goggles mounted source).” ’230 patent at 1:58-2:3.
It is evident that the Second Patent Family did not intend to disavow the First Family’s use
of orientation data, but rather sought to supplement the use of orientation data with an image check.
This is further evidenced by the remainder of the specification, which is replete with references to
sensors used to determine the orientation of the first and second video sources. See, e.g.,’230
patent at 1:64-66; 5:28-6:48; 8:21-10:15; 23:63-65; Fig. 3. Though other types of motion data,
such as position data (see, e.g., ’230 patent at 5:60-6:4), are referenced, the consistent reference to
orientation data in each embodiment provides a POSITA with reasonable certainty that motion
data includes at least orientation data.
Defendant argues that SAIC’s construction of “including orientation data” is not supported
by intrinsic data because “when the inventors sought to use ‘orientation data’ or even ’position
data’ they did so expressly.” Government’s Opening Cl. Constr. Br. at 17-18. Therefore, the
Government contends the inventors’ omission of orientation data from the claims indicates that
65
the inventors intended to use motion data differently than SAIC proposes. Id. Additionally, the
Government cites to a preferred embodiment which explains that ultra-wideband radios can be
used rather than using separate orientation sensors. Markman Hearing Tr. at 139:22-140:24. Thus,
according to the Government, “at least orientation data” would exclude this embodiment.
These arguments miss the point. SAIC is not arguing that orientation equals motion data.
Motion data may include position data. See, e.g., ’230 patent at 8:21-28 (“After initial calibration,
computer 30 receives position data for system 10 from GPS chipset 31 (FIG. 2) and/or from data
communication chipset 32 in block 105.”). However, SAIC’s construction simply requires that
orientation data be used. The embodiment’s ultra-wideband radios section, cited by the
Government, is not to the contrary. The specification consistently states that devices other than
IMUs can be used to determine orientation. See, e.g., ’230 patent at 4:17-41. The embodiment
does not state that orientation is not used; rather, it states that orientation data based on relative
alignment of ultra-wideband radios can be used instead of IMUs. Id. Accordingly, contrary to the
Government’s arguments, SAIC’s construction is consistent with the intrinsic evidence.
The Government relies on the testimony of, Drs. Neumann, Azuma, and Welch to support
its assertion that “motion data” is indefinite. As explained below, the cited extrinsic evidence is
consistent with SAIC’s claim construction and does not amount to clear and convincing evidence
of indefiniteness.
First, the Government relies on statements made by Dr. Azuma during his deposition that
“motion data” is “not . . . a well-defined term in the field.” Azuma Dep. at 126:20-127:12, 80:23-
81:20. This statement standing alone does not amount to clear and convincing evidence that
“motion data” as used in the patents is indefinite—primarily because, as the Government concedes,
Dr. Azuma did not read the patents prior to his deposition. Azuma Dep. at 238:8-25 (“I have not
66
read the SAIC patents. . . . I am not familiar with those particular patents.”). Moreover, when
asked how Dr. Azuma used “motion data,” Dr. Azuma’s explanation comports with SAIC’s
proposed usage of the term. Specifically, Dr. Azuma stated that “motion data” includes, inter alia,
orientation data. Azuma Dep. at 160:3-16. Thus, to the extent Dr. Azuma’s testimony is at all
relevant, it comports with SAIC’s construction of motion data.
Next, the Government relies on a statement made by Dr. Welch during his deposition that
SAIC’s proposed construction is certainly “not elegant” when applied to the claims. Joint
Submission of Cl. Constr. Experts, Ex. 2 Gregory Welch Deposition Transcript (Welch Dep.)
(ECF No. 79-2) at 156:7-17. Again, this statement is not clear and convincing evidence of
indefiniteness. Moreover, Dr. Welch ultimately opined that the asserted patents teach that sensor
data indicating the movement of weapon and goggle components (i.e., data indicative of motion)
includes orientation data and that this orientation data is used to determine a location of the weapon
image within the goggles image. See Pl.’s Opening Cl. Constr. Br., Ex. L Declaration of Gregory
Welch (Welch Decl.) (ECF No. 90-15) ¶ 63 (identifying patent disclosures in support of SAIC’s
construction of motion data).
Lastly, to support its position, the Government points to the following exchange between
Dr. Neumann and Plaintiff’s counsel during Dr. Neumann’s deposition:
Q. So the original question where you asked me to point you to some pieces of the
specification was would you agree that the specification of the ‘230 patent that we
had marked as Exhibit 8 discloses use of a rotation data to indicate relative motion
between the two – the two video sources.
A. Okay
Q. Would you agree with that?
A. The patent uses – the description uses specific terms like roll, pitch and yaw.
Okay. Motion data can be many things. There are not just a single roll, pitch and
yaw in this patent. There are multiples. There is a sub G. There’s the sub S.
67
There’s – there’s no sub anything. There’s just yaw and roll described. There are
so many different types of data described that I think it adds to the confusion of
what motion data means.
Q. Well – okay. So the sub G and the sub S refer – correlate to each of the video
sources, I believe. Would you agree with that?
A. They – as I recall, they deal with the gun motion and the S was –
Q. The scope, I believe. Although I would have to –
A. See, I’m not sure anymore. But yes. There’s two different things that are
moving. So we have two different suffixes.
Q. Okay.
A. Then there’s the data values from sensors indicate vertical rise pitch. So there’s
just – there’s so many different types of motion data mentioned, things that could
be motion data mentioned, which one – which one is it?
Q. Why couldn’t it be any of those? Just – why is that a problem for you?
…
A. If you pick the wrong data, you may not be able to accomplish your purpose.
It’s really important in technical documents to be clear. This signal goes from here
to here. You can’t just say a signal goes from here to here. It could be any signal.
Okay. When they say a motion data is used for this or for that, which data?
…
A. I mean, that the essence of – that’s the essence of why I say it’s indefinite. There
is no definition.
Joint Submission of Cl. Constr. Experts, Ex. 2 Ulrich Neumann Deposition Transcript (Neumann
Dep.) (ECF No. 97-1) at 219:10-221:9.
Dr. Neumann’s testimony does not amount to clear and convincing evidence that the term
“motion data” is indefinite. An expert may articulate the meaning of a term to a POSITA, but then
the court must conduct a legal analysis to see if that same meaning fits with the term “in the context
of the specific patent claim under review,” because “experts may be examined to explain terms of
art . . . but they cannot be used to prove the proper or legal construction of any instrument of
68
writing.” See Teva Pharms. USA, Inc. v. Sandoz, Inc., 574 U.S. 318, 331 (2015) (internal
quotations and citations omitted). Although Dr. Neumann expressed concern that “motion data”
may include data beyond orientation data, Dr. Neumann does not dispute that orientation data is
subsumed within the broader category of motion data. He simply concluded that there are “so
many different types of motion data mentioned [in the 230 patent’s specification, i]f you pick the
wrong data, you may not be able to accomplish your purpose.” See Neumann Dep. at 220:14-23.
It is well-established, however, that a term is not indefinite simply because it is broad. See BASF
Corp. v. Johnson Matthey Inc., 875 F.3d 1360, 1367 (Fed. Cir. 2017) (“But the inference of
indefiniteness simply from the scope finding is legally incorrect: ‘breadth is not indefiniteness.’”
(quoting SmithKline Beecham Corp. v. Apotex Corp., 403 F.3d 1331, 1341 (Fed. Cir. 2005))). As
Dr. Neumann acknowledges, the specification explicitly discloses multiple types of motion data.
Neumann Dep. at 219:10-221:9. Contrary to Dr. Neumann’s assertions, these extensive
disclosures are a source of guidance rather than confusion. A POSITA seeking to interpret the
bounds of motion data need simply refer to the patent document, which consistently refers to
“orientation data” about the first and second video source. Aside from the numerosity of examples
of motion data and a lack of an explicit definition, Dr. Neumann was unable to articulate how the
term motion data as used in the patent is indefinite. Instead, Dr. Neumann summarily concluded
that “[i]f you pick the wrong data, you may not be able to accomplish your purpose.” See id. at
221:22-23 (emphasis added). To prove indefiniteness by clear and convincing evidence, the
Government must do more. See Apple Inc. v. Samsung Elecs. Co., 786 F.3d 983, 1003 (Fed. Cir.
2015) (attempting to discredit the patentee’s experts is not sufficient to find claim indefinite), rev’d
on separate grounds, 137 S. Ct. 429 (2016), remanded to 678 Fed. App’x 1012 (2017); Microsoft
Corp. v. i4i Ltd. P’ship, 564 U.S. 91, 104-05 (2011) (referencing many instances where the Court
69
has required a heightened standard of proof to overcome patent’s presumption of validity).
In sum, Drs. Neumann, Azuma, and Welch were each able to identify orientation data as a
component of “motion data.” Accordingly, for the reasons stated above, the Court finds that the
term “motion data” is definite, and its proper construction is “data indicative of motion, including
at least orientation data.”
VI. “IN REGISTRATION WITH” / “REGISTERING”
SAIC Government Microsoft L3
“in proper alignment Indefinite N/A Indefinite
and
position, so as to
coincide
and not be
substantially offset” 18
Court’s Construction: Indefinite
L3 and the Government argue that the registration terms “in registration with” and
“registering” are indefinite because the terms are (1) subjective and the patents fail to disclose
parameters for acceptable degrees of error or inform a POSITA when registration is achieved, and
(2) the patents fail to disclose how registration is accomplished. Government’s Opening Cl.
Constr. Br. at 39-40; L3’s Opening Cl. Constr. Br. at 22. The Government submits that, without
such criteria, a POSITA reviewing the intrinsic record would not be able to determine the objective
scope of the registration terms with reasonable certainty. Government’s Responsive Cl. Constr.
Br. at 24-25.
18
In the parties’ Joint Claim Construction Statement, SAIC proposed the following construction:
70
SAIC argues that “registration” means “in proper alignment and position, so as to coincide
and not be substantially offset.” Pl.’s Opening Cl. Constr. Br. at 20, 30. SAIC further contends
that “registration” is definite because the patents disclose examples of what is and what is not
proper alignment to inform when registration is achieved. Pl.’s Opening Cl. Constr. Br. at 21-23
(citing ’012 patent at 2:15-18, 3:24-4:6, 6:25-27, 6:38-7:12, Figs. 1, 4, and 8; ’230 patent at 10:11-
15; ’012 Patent File History at 10-11). Additionally, the patents disclose exemplary methods for
using orientation data to show how registration can be accomplished. Pl.’s Opening Cl. Constr.
Br. at 21, 23 (citing ’012 patent at Abstract, 2:30-38, 4:51-54, 6:38-45, 7:30-42, 8:5-48, Figs. 8,
9A-12B; ’230 patent at Abstract, 3:46-50, 3:67-4:16, Figs. 6A-H).
This Court agrees with the Government and L3. Pursuant to statute, a specification must
“conclude with one or more claims particularly pointing out and distinctly claiming the subject
matter which the inventor or a joint inventor regards as the invention.” 35 U.S.C. § 112(b). A
claim fails to satisfy this statutory requirement and is thus invalid for indefiniteness if its language,
when read in light of the specification and the prosecution history, “fail[s] to inform, with
reasonable certainty, those skilled in the art about the scope of the invention.” Nautilus, 572 U.S.
at 901.
Here, the patents fail to provide objective criteria for a POSITA to determine with
reasonable certainty when registration is accomplished. The intrinsic evidence does not contain
any criteria or other description by which to measure or know when “wherein boundaries of the
“in proper alignment and position, so as to coincide and not be substantially offset.” Joint Cl.
Constr. Statement, Ex. 1 Joint Cl. Constr. Chart (ECF No. 63-1) at 8. In its opening claim
construction brief, SAIC substituted “substantially” with “distinctly” so that its proposed
construction reads, “in proper alignment and position, so as to coincide and not be substantially
(i.e., distinctly) offset. Pl.’s Opening Cl. Constr. Br. at 20. At oral argument, the Government
stated that it believed that this change should not affect this Court’s indefiniteness analysis.
Markman Hearing Tr. at 230:23-231:7.
71
displayed images are in registration with boundaries of portions of the visual field represented by
the displayed images.” See Berkheimer v. HP Inc., 881 F.3d 1360, 1364 (Fed. Cir. 2017) (finding
the term “minimal redundancy” indefinite because the patent lacked objective boundaries). The
First Patent Family does not define registration. Instead, it uses examples to disclose when
registration is accomplished. For example, ’012 Patent Figure 4 depicts an “image produced by
an illustrative embodiment of the invention,” in which images from a weapon sight feed are in
proper position and alignment (i.e., coincide and are not distinctly offset) with images seen via a
HUD. ’012 patent at 3:56-57, Fig. 4. Conversely, in describing the prior art system of Figure 1,
the specification discloses that “[b]oth images depict the same subjects, a group of soldiers
accompanying an armored personnel carrier (APC)[,]” but “[t]he two images are distinctly offset,
with . . . the same target appearing in different places in the field of view . . . .” Id. at 2:9-18.
The Second Patent Family incorporates the First Family’s disclosure and consistently
explains that “‘registration’ refers to positioning of a scope image (or portion of that scope image)
within a goggles image so that the two images are properly aligned and positioned, and one image
coincides with the other.” ’230 patent at 10:11-15.
Here, neither SAIC’s cited examples, nor its proposed definition adequately inform a
POSITA of the invention’s metes and bounds with reasonable certainty. Registration depends on
the perspective of a particular application or user, the method of registration used, and the needs
and precision required by the particular use in which the user is engaged. See Neumann Dep. at
93:7-15; Welch Dep. at 203:22-212:16. At his deposition, Dr. Neumann explained that “in a
nutshell” the field of registration involves computations designed to measure and address
registration errors. Neumann Dep. at 93:22-94:7. In essence, Dr. Neumann testified, “registration”
is a term of degree in that it is context-dependent and measurable. See Neumann Dep. at 86-87:4,
72
120:25-121:10. While terms of degree are not inherently indefinite, the patent must provide some
objective criteria for a POSITA to determine the scope of the invention with reasonable certainty.
Interval Licensing LLC, 766 F.3d at 1370-71 (citations omitted).
Neither the ’230 patent’s use of the terms “proper alignment” or “distinctly offset,” nor
SAIC’s proposed construction of “substantially offset” provide a POSITA with objective criteria
to determine whether “registration” is achieved. As explained by both SAIC’s and the
Government’s experts, there are a wealth of registration techniques with different variations and
parameters. Neumann Dep. at 120:25-129:16; see also Welch Dep. at 203:22-212:16. What is
considered proper alignment using one measurement technique will not be considered proper
alignment using another technique. See Neumann Dep. at 120:25-129:16; Welch Dep. at 203:22-
212:16.
The First Patent Family does not disclose any registration technique or combination of
techniques for measuring whether registration has been accomplished. Rather the patents rely on
the high-level flow chart in Figure 8 and direct a POSITA to registration techniques that “are well
known in the art.” See, e.g., ’012 patent at 9:36-52 (“well known rigid or non-rigid image
registration techniques . . . to register the images by, for example, finding common visual elements
between them.”); ’012 patent at 6:62-64 (“Various algorithms for rotating an image by a certain
number of degrees are well known in the art.”).
The Second Patent Family includes similarly high-level flow charts. The flow chart in
Figures 5A and 5B explain the registration process for the Second Patent Family. ’230 patent at
7:46-11:29. As noted, registration is initially accomplished using motion data which is defined
broadly as “including at least orientation data.” See supra Section V: Motion Data. The alignment
accomplished with motion data is then “evaluated” using image data and adjusted as needed before
73
it is displayed on the HUD. Id. at 11:30-49. The written description states that “[t]he steps shown
in FIGS. 5A and 5B (and in other flow charts described below) can be reordered, combined, split,
replaced, etc.” Id. at 7:46-49. While the Second Patent Family references some of the metrics
traditionally used to measure registration, these metrics must have objective bounds, and nothing
in the patent explains when certain metrics would be used over others. See, e.g., ’230 patent at
10:48-11:30 (discussing the use of Brouwer’s fixed point to check positioning of the images); see
also Neumann Decl. ¶¶ 65-66. The Second Patent Family also implies that additional unnamed
metrics may be used. For example, the ’230 patent mentions that peak to sidelobe ratio (PSR) is
one metric to determine registration, but also mentions that “numerous definitions of PSR are
known in the art . . . .” ’230 patent at 11:11-17. However, the ’230 patent does not reference PSR
or any other objective metric when defining “registration.”
These high-level flow charts give very little guidance to a POSITA as to how registration
is objectively measured. This lack of guidance prevents a POSITA from ascertaining the scope of
registration with reasonable certainty because different registration methodologies involve
different parameters for determining whether registration has been accomplished. Ball Metal
Beverage Container Corp. v. Crown Packaging Tech., Inc., 838 F. App’x 538, 542-43 (Fed. Cir.
2020) (“Under our case law, then, a claim may be invalid as indefinite when (1) different known
methods exist for calculating a claimed parameter, (2) nothing in the record suggests using one
method in particular, and (3) application of the different methods result in materially different
outcomes for the claim's scope such that a product or method may infringe the claim under one
method but not infringe when employing another method. Such a claim lacks the required degree
of precision ‘to afford clear notice of what is claimed, thereby apprising the public of what is still
open to them.’” (quoting Nautilus, 572 U.S. at 909)). Dr. Neumann opined that there are multiple
74
methods of “evaluation of registration error and a minimization process.” Neumann Decl. ¶ 54.
“For example, cross-correlation measures, Fourier phase correlation, and point mapping are
common registration metrics.” Id. Dr. Welch also agreed that different image registration
techniques will lead to different results. Specifically, Dr. Welch testified that:
Q. Well, it’s fair to say if you take the same pair of images and you register them
using different transformation, you might get different results. Fair?
A. Well, as I said earlier, a lot of different things will affect whether you get -- the
results would be different or not. In fact, it seems highly unlikely that any two --
you know, it would be very small things that would vary that would give you
different results.
The transformations that I see here are no different than I think the
transformations, at least some of these that were talked about in, I think the ‘012
patent where it said that these were common, when it was talking about rotations,
image rotations. These are the sort of things I think is what was meant there.
Q. Maybe I missed it, but so it’s fair to say that using different transformations,
you might get different results . . . for registration?
A. You would get different results, and as I said earlier, I think -- I think you would
likely get different results. As I said earlier, the differences would -- the impact of
those differences or the importance of those differences would depend on the use
case, the -- you know, the people who are developing the system.
Welch Dep. at 211:9-212:16.
In sum, testimony from both experts establishes that registration can be measured multiple
ways and that different measuring techniques will yield different results in determining whether
registration has occurred. Both experts also agreed that determining whether registration is
achieved is dependent upon the application and the user’s tolerance for registration errors. Nothing
in the patent itself provides a POSITA with any objective measure as to the bounds of registration.
Without any objective measure, a POSITA is left “to consult the unpredictable vagaries of any one
person’s opinion” to determine whether registration occurred. Dow Chem. Co. v. Nova Chemicals
Corp. (Canada), 803 F.3d 620, 635 (Fed. Cir. 2015) (internal quotations and citation omitted)
75
(holding term indefinite where invention did not disclose a method to determine whether a claim
parameter was met).
Notwithstanding the patents’ failure to provide an objective metric to this highly contextual
term, SAIC argues that the patents provide reasonable certainty by disclosing examples of a
registered image and an unregistered image. Pl.’s Opening Cl. Constr. Br. at 21.
This argument is without merit. While SAIC is correct in that a claim term can be rendered
definite through the use of examples that provide points of comparison, those examples must
provide some objective criteria by which a POSITA can determine the scope of a claim with
reasonable certainty. Sonix Tech. Co., Ltd. v. Publ’ns Int’l, Ltd., 844 F.3d 1370, 1379 (Fed. Cir.
2017) (holding “visually negligible” definite because examples in the specification provided
objective criteria by which a POSITA could identify the scope of the invention with reasonable
certainty); Guangdong Alison Hi-Tech Co. v. Int’l Trade Comm’n, 936 F.3d 1353, 1360–62 (Fed.
Cir. 2019) (holding the term “lofty fibrous batting” definite where the specification provided seven
detailed examples for comparison, and the parties’ expert testimony supported the conclusion that
a POSITA could objectively identify characteristics of the term); One-E-Way, Inc. v. Int’l Trade
Comm’n, 859 F.3d 1059, 1066 (Fed. Cir. 2017) (finding “virtually free from interference” definite
where statements in the specification and prosecution history indicated that the phrase meant “free
from eavesdropping,” which provided an objective standard to inform a POSITA of the scope of
the invention with reasonable certainty). Indeed, the mere existence of examples in the written
description will not always render a claim definite. Sonix Tech., 844 F.3d at 1380.
As explained above, “registration” is understood on a continuum. See Neumann Dep. at
86:21-87:4, 120:25-124:3, 175:9-176:4, 181:5-8, 193:12-16. Here, the examples provided in the
patents at issue do not provide objective criteria to inform a POSITA of the scope of registration.
76
The parties agree the patents do not claim perfect registration. See Welch Dep. at 165:24-166:25
(“There is no perfect registration.”), 167:5-168:10 (“[P]erfect registration doesn’t exist[,]” even in
the context of the asserted patents), 169:8-25 (“[T]here’s no system that is going to match every
pixel in intensity pixel for pixel.”). Outside of perfect registration, the concept of registration here
is context dependent. Welch Dep. at 203:22-204:6. Because registration depends on context, there
are no inherent objective parameters that a POSITA can use to determine the scope of the term.
See Interval Licensing LLC, 766 F.3d at 1371-74. The patent, therefore, needs to provide some
objective criteria for assessing an acceptable variance from perfect registration for a POSITA to
determine the scope of the registration terms with reasonable certainty. Id.
The First Patent Family provides two examples, which are incorporated by reference in the
Second Patent Family. ’012 patent at Figs. 1, 4. In Figure 4 of the First Patent Family, the patents
state that the images in that figure are registered but do not disclose any offset. See Welch Dep.
at 165:24-166:25. In the counterexample taken from the prior art, Figure 1 illustrates a very
significant offset, which the patents describe as not in registration. ’012 patent at 2:4-28. The
tremendous gap between these two examples creates a zone of uncertainty. See Nautilus, 572 U.S.
at 909-10. With examples only providing boundaries at the extremes, a skilled artisan is left to
wonder what other images could fall between these two figures and still be considered registered.
See, e.g., Automated Pack’g Systems, Inc. v. Free Flow Pack’g Int’l, Inc., No. 18-CV-00356-EMC,
2018 WL 3659014, at *18 (N.D. Cal. Aug. 2, 2018) (finding examples of an inserting device that
could not be as small as a needle, nor as big as a baseball bat, were too extreme to provide a
POSITA with objective criteria to determine the scope of the patent claims); Power Integrations,
Inc. v. ON Semiconductor Corp., No. 16-CV-06371-BLF, 2018 WL 5603631, at *20 (N.D. Cal.
Oct. 26, 2018) (holding the term “moderate power threshold value” indefinite where the patent
77
provided an extreme range of very low values to the maximum peak value).
The patents’ disclosure of registration errors also does not provide the required “reasonable
certainty.” See Nautilus, 572 U.S. at 901. The patent specifications discuss parallax as a possible
source of registration error. See, e.g., ’012 patent at 8:49-9:35. Specifically, the ’012 patent
mentions parallax “error of about 2.9 degrees in the placement of the video frame” when the target
is at 10 meters and “is to some extent a non-issue [as t]he system proposed would likely be used
for targets greater than 10 meters more often than not.” Id. at 8:60-67. In its brief, SAIC contends
that the ’012 patent—and a similar discussion exists for the ’230 patent—describing an error of
2.9 degrees at 10 meters and then fewer degrees at greater target distances is an example of criteria
from which a POSITA can assess an acceptable degree of error, e.g., 2.9 degrees or less in the
video frame. Pl.’s Opening Cl. Constr. Br. at 27; see also ’012 patent at 8:49-64, ’230 patent at
9:31-47. The problem with this argument is that the patent described an error of 2.9 degrees at 10
meters assuming the scope image and goggle image were “perfectly aligned,” ’012 patent at 8:52-
53, which the parties’ experts acknowledge is impossible. See Welch Dep. at 166:8 (“There is no
perfect registration.”); Neumann Dep. at 186:24-187:11 (“[I]t’s fair to say there will be deviation,
there will be error [in registration.]”). Thus, a POSITA could not rely on the disclosure of “2.9
degrees or less” as an objective criterion for when registration is achieved. See Neumann Decl.
(ECF No. 66) ¶ 57 (explaining why the disclosure of parallax error does not provide “objective
boundaries and a POSITA would not be able to discern [the registration terms] with reasonable
certainty”).
The lack of guidance provided by the examples in the patents is confirmed by the extrinsic
evidence. Both SAIC’s and the Government’s experts were unable to identify objective criteria
for determining registration. Dr. Neumann opined, “[t]he intrinsic evidence does not contain any
78
criteria or other description by which to measure or know when [the claim term] ‘wherein the
boundaries of the displayed images are in registration with boundaries of portions of the visual
field represented by the displayed images.’” Neumann Decl. ¶ 53 (emphasis omitted). When
discussing the patents’ disclosure of registration errors, Dr. Neumann also opined that the
registration of overlapping images is impacted by several issues, such as “photonic (intensity)
variations and image acquisition . . . differences[,]” not just parallax. Neumann Decl. ¶ 54.
When SAIC's expert, Dr. Welch, was asked whether the patents provide any basis for
determining what level of offset is acceptable, he stated, “[s]itting here right now, I don't recall
that there’s anywhere where the patents disclose something like acceptable offsets.” Welch Dep.
at 170:2-18. To put this comment in context, Dr. Welch came to this conclusion directly following
a question about Figure 4 of the ’012 patent. Specifically, Dr. Welch was asked “[s]o you can't
estimate whether there's any offset between the two images that are allegedly registered in Figure
4 of the ’012 patent; correct?” Welch Dep. at 169:8-11. Dr. Welch responded:
A: I as a human would have to understand what we mean by offset, because there
are who knows how many pixels in that inset region. There's a lot of information
in the middle, and we'd have -- and around the boundaries, and we'd have to sort of
agree on what's important or not important because there's no system that is going
to match every pixel in intensity pixel for pixel. So we’d have to agree on what it
means to be okay. What I'm saying is the patent teaches that this is what they mean
by registration.
Welch Dep. at 169:13-25. The exchange demonstrates that Dr. Welch does not view Figures 1
and 4 as objective guidance to inform a POSITA of the objective boundaries of registration. Dr.
Welch does not specifically reference the patents’ disclosure of 2.9 degrees or less as an objective
criterion for when registration is achieved in either his declaration or his deposition. In fact, in his
declaration, Dr. Welch asserts that examples for addressing parallax and calibration methods are
unnecessary for a POSITA to understand the patents’ claims with reasonable certainty. Welch
79
Decl. ¶ 57. This assertion is inconsistent with Federal Circuit precedent.
The Federal Circuit’s decision in Berkheimer v. HP Inc., 881 F.3d 1360 (Fed. Cir. 2017) is
instructive. In Berkheimer, the Federal Circuit held the use of “minimal redundancy” to be an
indefinite term of degree for which there must be objective boundaries. Id. at 1364. The patent in
Berkheimer related to digitally processing and achieving files in a digital asset management
system, which eliminated redundancy of common text and graphical elements. Id. at 1362.
Relying on intrinsic evidence as well as an expert declaration, the district court had found that
minimal redundancy was “highly subjective” and would not provide a POSITA with objective
criteria for when minimal redundancy was achieved. Id. at 1363. The Federal Circuit agreed. Id.
The Federal Circuit first noted that the specification inconsistently described the invention. Id. at
1363-64. In certain places, the specification described the system as “minimizing redundant
objects,” while elsewhere it stated that redundancy was “eliminat[ed].” Id. at 1364. In the
prosecution history, the inventor had stated that the claim at issue desired to eliminate redundancy
but used the term “minimal” because eliminating redundancy was unlikely. Id. Moreover, the
only example included in the specification exhibited no redundancy. Id. Accordingly, the Federal
Circuit stated that “[t]he specification contains no point of comparison for skilled artisans to
determine an objective boundary of ‘minimal’ when the archive includes some redundancies.” Id.
(citation and emphasis omitted). The Federal Circuit held that terms of degree require objective
boundaries to inform a POSITA of the scope of the invention, and no objective boundary was
provided where the invention failed to inform “how much is minimal.” Id. (emphasis in original).
Here, like in Berkheimer, neither the disclosure of examples nor registration errors provide
reasonable certainty of the bounds of registration. As was the case in Berkheimer, because perfect
registration cannot be accomplished, the patents must disclose an acceptable level of “offset.” The
80
examples here are too extreme to provide objective criteria to a POSITA; and, like Berkheimer,
expert testimony supports this conclusion. Both experts agree that (1) registration is context
dependent, (2) the patents do not claim perfect registration, and (3) the patents do not disclose an
acceptable offset. Accordingly, for the reasons stated above, the Court holds the terms “in
registration with” and “registering” to be indefinite, as no objective way exists to calculate how
much offset is acceptable.
*****
81
CONCLUSION
For the foregoing reasons, the Court construes:
1. “video images” / “video source image” / “video data of images” as “digital, analog,
or nonstandard video frames;”
2. “transparent display” as “a display that has the effect of being transparent or
translucent, allowing simultaneous viewing of the underlying visual field and other
images or information;”
3. “overlay” in accordance with its plain and ordinary meaning, “e.g., Overlaying:
positioned over or upon. Overlay/overlays: are over or upon/is over or upon;”
4. “based on a comparison” as “based on a comparison of image data (e.g., content and
contrast) from the first and second video source images;” and
5. “motion data” as “data indicative of motion, including at least orientation data.”
Additionally, the Court holds the terms “in registration with” / “registering” indefinite.
The parties are DIRECTED to file a Joint Status Report by August 23, 2021, proposing a
schedule for further proceedings.
IT IS SO ORDERED.
s/ Eleni M. Roumel
ELENI M. ROUMEL
Judge
Dated: August 6, 2021
Washington, D.C.
82