Fujifilm Corporation v. Motorola Mobility Holdings, Inc. et al

Filing 66

CLAIM CONSTRUCTION ORDER. The Court will hold a Case Management Conference on 12/17/2013 at 02:00 PM in Courtroom 2, 17th Floor, San Francisco, at which the parties should be prepared to discuss case scheduling and other matters that may require the Court's attention. Signed by Judge William H. Orrick on 11/18/2013. (jmdS, COURT STAFF) (Filed on 11/18/2013)

1
2
3
4

UNITED STATES DISTRICT COURT

5

NORTHERN DISTRICT OF CALIFORNIA

6
7

FUJIFILM CORPORATION,

Case No. 12-cv-03587-WHO

Plaintiff,
8
v.

CLAIM CONSTRUCTION ORDER

9
10

MOTOROLA MOBILITY LLC,
Defendant.

United States District Court
Northern District of California

11
12
INTRODUCTION

13
14

Fujifilm Corporation (―Fujifilm‖) brings this suit against Motorola Mobility LLC

15

(―Motorola‖) charging infringement of five patents: United States Patent Nos. 5,734,427 (―the

16

‗427 Patent‖); 7,327,886 (―the ‗886 Patent‖); 6,144,763 (―the ‗763 Patent‖); 6,915,119 (―the ‗119

17

Patent‖); and 8,306,285 (―the ‗285 Patent‖). The patents at issue relate to technology in cameras

18

and cellular phones.

19

On October 28, 2013, the Court held a Markman hearing on the construction of disputed

20

terms. Based on the parties‘ briefs and argument of counsel, the Court construes the disputed

21

terms as below.

22
23

LEGAL STANDARD
Claim construction is a matter of law. See Markman v. Westview Instruments, Inc., 517

24

U.S. 370, 372 (1996); Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582 (Fed. Cir. 1996).

25

Terms contained in claims are ―generally given their ordinary and customary meaning.‖ Vitronics,

26

90 F.3d at 1582. In determining the proper construction of a claim, a court begins with the

27

intrinsic evidence of record, consisting of the claim language, the patent specification, and, if in

28

evidence, the prosecution history. Phillips v. AWH Corp., 415 F.3d 1303, 1313 (Fed. Cir. 2005);

1

see also Vitronics, 90 F.3d at 1582. ―A claim term used in multiple claims should be construed

2

consistently . . . .‖ Inverness Med. Switzerland GmbH v. Princeton Biomeditech Corp., 309 F.3d

3

1365, 1371 (Fed. Cir. 2002).

4

―The appropriate starting point [ ] is always with the language of the asserted claim itself.‖

5

Comark Commc’ns, Inc. v. Harris Corp., 156 F.3d 1182, 1186 (Fed. Cir. 1998). ―[T]he ordinary

6

and customary meaning of a claim term is the meaning that the term would have to a person of

7

ordinary skill in the art in question at the time of the invention, i.e., as of the effective filing date

8

of the patent application.‖ Phillips, 415 F.3d at 1312. ―There are only two exceptions to this

9

general rule: 1) when a patentee sets out a definition and acts as his own lexicographer, or 2) when
the patentee disavows the full scope of a claim term either in the specification or during

11
United States District Court
Northern District of California

10

prosecution.‖ Thorner v. Sony Computer Entm’t Am. LLC, 669 F.3d 1362, 1365 (Fed. Cir. 2012).

12

―Importantly, the person of ordinary skill in the art is deemed to read the claim term not

13

only in the context of the particular claim in which the disputed term appears, but in the context of

14

the entire patent, including the specification.‖ Phillips, 415 F.3d at 1313. ―Claims speak to those

15

skilled in the art,‖ but ―[w]hen the meaning of words in a claim is in dispute, the specification and

16

prosecution history can provide relevant information about the scope and meaning of the claim.‖

17

Electro Med. Sys., S.A. v. Cooper Life Scis., Inc., 34 F.3d 1048, 1054 (Fed. Cir. 1994) (citations

18

omitted). ―[T]he specification is always highly relevant to the claim construction analysis.

19

Usually, it is dispositive; it is the single best guide to the meaning of a disputed term.‖ Vitronics,

20

90 F.3d at 1582. ―However, claims are not to be interpreted by adding limitations appearing only

21

in the specification.‖ Id. ―Thus, although the specifications may well indicate that certain

22

embodiments are preferred, particular embodiments appearing in a specification will not be read

23

into the claims when the claim language is broader than such embodiments.‖ Id. ―[T]he

24

description may act as a sort of dictionary, which explains the invention and may define terms

25

used in the claims,‖ and the ―patentee is free to be his own lexicographer,‖ but ―any special

26

definition given to a word must be clearly defined in the specification.‖ Markman, 517 U.S. at

27

989-90.

28

On the other hand, it is a fundamental rule that ―claims must be construed so as to be
2

1

consistent with the specification.‖ Phillips, 415 F.3d at 1316. ―The construction that stays true to

2

the claim language and most naturally aligns with the patent‘s description of the invention will be,

3

in the end, the correct construction.‖ Renishaw PLC v. Marposs Societa’ per Azioni, 158 F.3d

4

1243, 1250 (Fed. Cir. 1998).
Finally, the Court may consider the prosecution history of the patent, if in evidence.

5
6

Markman, 52 F.3d at 980. The prosecution history may ―inform the meaning of the claim

7

language by demonstrating how the inventor understood the invention and whether the inventor

8

limited the invention in the course of prosecution, making the claim scope narrower than it would

9

otherwise be.‖ Phillips, 415 F.3d at 1317 (citing Vitronics, 90 F.3d at 1582-83); see also Chimie
v. PPG Indus., Inc., 402 F.3d 1371, 1384 (Fed. Cir. 2005) (―The purpose of consulting the

11
United States District Court
Northern District of California

10

prosecution history in construing a claim is to exclude any interpretation that was disclaimed

12

during prosecution.‖) (internal quotations omitted).
In most situations, analysis of this intrinsic evidence alone will resolve claim construction

13
14

disputes. Vitronics, 90 F.3d at 1583. However, ―it is entirely appropriate . . . for a court to consult

15

trustworthy extrinsic evidence to ensure that the claim construction it is tending to from the patent

16

file is not inconsistent with clearly expressed, plainly apposite, and widely held understandings in

17

the pertinent technical field.‖ Pitney Bowes, Inc. v. Hewlett-Packard Co., 182 F.3d 1298, 1309

18

(Fed. Cir. 1999). Extrinsic evidence ―consists of all evidence external to the patent and

19

prosecution history, including expert and inventor testimony, dictionaries, and learned treatises.‖

20

Markman, 52 F.3d at 980. All extrinsic evidence should be evaluated in light of the intrinsic

21

evidence, Phillips, 415 F.3d at 1319, and courts should not rely on extrinsic evidence in claim

22

construction to contradict the meaning of claims discernible from examination of the claims, the

23

written description, and the prosecution history, Pitney Bowes, 182 F.3d at 1308 (citing Vitronics,

24

90 F.3d at 1583). While extrinsic evidence may guide the meaning of a claim term, such evidence

25

is less reliable than intrinsic evidence. Phillips, 415 F.3d at 1318-19.
DISCUSSION

26
27
28

I.

PATENT NO. 5,734,427
This patent relates to an electronic still camera that displays on a monitor in real-time the
3

1

subject upon which the camera is focusing. ‗427 Patent 1:12-14. It can also record images of the

2

subject. The invention allows for a high-resolution image to be processed, or ―thinned,‖ into a

3

low-resolution image. ‗427 Patent 1:60-65.

4

A. “thinning”

5

The first disputed term is found in Claims 1, 6, 11, 14, 15, and 19.

6
7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18
19
20
21

Disputed Term
Claim 1:
Fujifilm: ―thinning said
first image signal to
thereby produce a second
image signal having a low
resolution‖
Motorola: ―thinning‖
(―said first image signal
having a high resolution‖)
Claim 6:
Fujifilm: ―thinning said
high-resolution pixel data
output from said imaging
device to thereby produce
low-resolution image
data‖

Fujifilm’s Constructions1
reducing the resolution of the
first image signal to create a low
resolution image signal
representative of the subject

Motorola’s Constructions2
reducing the resolution of the
[image data] before any color
interpolation into simultaneous
RGB image data

reducing the resolution of
the pixel data output from the
imaging device to create lowresolution image data
representative of the subject

reducing the resolution of the
[image data] before any color
interpolation into simultaneous
RGB image data

Motorola: ―thinning‖
(―said high-resolution
pixel data output from said
imaging device‖)
Claim 11: ―thinning said
reducing the resolution of the
pixel data‖
pixel data

22
23

Claims 14 & 15:

24

Fujifilm: ―thinning said
high-resolution image data
output from said imaging

25

reducing the resolution of the
image data output from the
imaging device to create low
resolution image data
representative of the subject

reducing the resolution of the
[image data] before any color
interpolation into simultaneous
RGB image data
reducing the resolution of the
[image data] before any color
interpolation into simultaneous
RGB image data

26
1

27
28

Fujifilm states that it has no objection to replacing ―pixel resolution‖ in its original constructions
with ―resolution.‖ Reply 1.
2
Motorola‘s construction previously contained the phrase ―using weighted averaging,‖ but it has
now removed it. Response 20 n.6.
4

1
2
3
4
5
6
7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18
19
20
21

22
23
24
25
26
27

device to thereby produce
low-resolution image
data‖
Motorola: ―thinning‖
(―said high-resolution
image data output from
said imaging device‖)
Claim 19:
Fujifilm: ―thinning said
first image signal . . . to
thereby produce a second
image signal having low
resolution‖

reducing the resolution of the
image signal to create a low
resolution image signal
representative of the subject

reducing the resolution of the
[image data] before any color
interpolation into simultaneous
RGB image data

Motorola: ―thinning‖
(―said first image signal‖)
The parties dispute whether thinning must occur before color interpolation into
simultaneous RGB data. Fujifilm‘s Br. (Dkt. No. 45) (―Br.‖) 3; Motorola‘s Response (Dkt. No.
48) (―Response‖) 21.
The Court adopts Fujifilm‘s constructions and finds that thinning need not occur before
color interpolation into simultaneous RGB image data. The claims at issue do not show that
thinning necessarily occurs before interpolation, as Motorola proposes. Claims 1 and 6 discuss
―thinning‖ but do not discuss ―interpolating‖ at all. Claims 11, 14, 15, and 19 discuss both.
Claims 11, 14, and 15 all identically disclose ―a thinning circuit for receiving [ ] pixel data . . . and
for thinning said pixel data,‖ after which they identically disclose ―an interpolating circuit for
receiving said pixel data from said thinning circuit.‖ ‗427 Patent 13:44-52; 14:40-48; 15:35-43.
Claim 19 similarly discloses a method of ―thinning [a] first image signal to thereby produce a
second image signal,‖ then ―interpolating from said second image signal.‖ ‗427 Patent 17:24-30.
The fact that the inventor did not discuss interpolation at all in Claims 1 and 6 strongly suggests
that a limitation in Claims 11, 14, 15 and 19 should not be read into the first two claims. See
Phillips, 415 F.3d at 1314 (―Differences among claims can also be a useful guide in understanding
the meaning of particular claim terms.‖).
Nothing else in the claim language mandates an order to these operations, and Motorola

28
5

1

provides no argument that it does. Further, Motorola points to no evidence in the patent that the

2

invention could not be realized unless its construction was adopted. Motorola‘s construction

3

unnecessarily inserts a limitation in some claims into other claims.
Motorola cites Biogen Idec, Inc. v. Glaxosmithkline LLC for the proposition that ―when the

4
5

patentee unequivocally and unambiguously disavows a certain meaning to obtain a patent, the

6

doctrine of prosecution history disclaimer narrows the meaning of the claim consistent with the

7

scope of the claim surrendered.‖ 713 F.3d 1090, 1095 (Fed. Cir. 2013) (emphasis added). It

8

argues that thinning must occur before interpolation because the patentee said so. Citing to the

9

prosecution history, it contends that the ―Remarks‖ by the applicant to the patent examiner, in a
section entitled ―Synopsis of the Present Invention,‖ constitutes a waiver: ―[T]he data

11
United States District Court
Northern District of California

10

representative of a high resolution image are thinned and then interpolated. More specifically,

12

high resolution image data . . . captured by the image sensor are thinned into image data of the

13

standard television format . . . . The thinned image data are then interpolated by the weighting

14

average means to be converted into simultaneous RGB image data, in which each pixel of data

15

includes R, G and B component data . . . .‖ Cangro Decl. Ex. 6 at 12.
The ‗427 Patent applicant‘s statements are neither an unequivocal nor unambiguous

16
17

disavowal of anything. Reading the statement in context, there is no indication that the applicant

18

was making a disclaimer in order to overcome an examinee‘s objection, nor is it clear that the

19

applicant was intentionally limiting the scope of the patent. As Biogen holds, a prosecution

20

disavowal must be ―clear and unmistakable‖ to overcome the ―heavy presumption‖ that claims

21

mean what they say. Id. The applicant‘s statements must be clearer than the examples provided

22

by Motorola for them to constitute a disavowal or estoppel. Motorola‘s reliance on the

23

prosecution history to support its construction is unavailing.3

24

B. “a viewfinder . . . implemented by a video monitor”

25

The second disputed term is found in Claims 1, 6, 14, and 15.

26
3

27
28

The parties dispute whether it is literally possible for interpolation to occur before thinning. The
Court does not have sufficient information before it to address this issue. In any event, the Court
need not decide that issue because it finds that there is enough evidence to support Fujifilm‘s
construction.
6

1
2
3

Fujifilm’s Construction
Plain and ordinary meaning;
Alternatively: a monitor serving as a
viewfinder

Motorola’s Construction
a viewfinder […] implemented by a video
monitor physically separate from the
camera body

4

The invention converts image data into a video signal, which is then displayed on a

5

monitor. The monitor serves as a viewfinder which allows the user to frame an image to be stored

6

in memory. Br. 5. The parties dispute whether the video monitor must be built into the camera.

7

The Court finds that this term needs no construction and that the ―viewfinder‖ referenced
in the term is not limited to ―video monitor[s] physically separate from the camera body,‖ as

9

Motorola‘s proposed construction states. Reviewing the claims at issue, there is no evidence that a

10

person of ordinary skill in the art would conclude that the ―video‖ monitor needs to be ―physically

11
United States District Court
Northern District of California

8

separate from the camera body.‖ The claims do not distinguish between built-in and physically

12

separate monitors, nor do they discuss the issue at all. The same is true of the written description.

13

In addition, the Court finds that a juror would understand the term based on its plain and ordinary

14

meaning.

15

The Federal Circuit instructed that ―[g]eneral descriptive terms will ordinarily be given

16

their full meaning; modifiers will not be added to broad terms standing alone.‖ Johnson

17

Worldwide Assocs., Inc. v. Zebco Corp., 175 F.3d 985, 989 (Fed. Cir. 1999) (citing Va. Panel

18

Corp. v. MAC Panel Co., 133 F.3d 860, 865-66 (Fed. Cir. 1997) (unmodified term ―reciprocating‖

19

not limited to linear reciprocation), and Bell Commc’ns Research, Inc. v. Vitalink Commcn’s

20

Corp., 55 F.3d 615, 621-22 (Fed. Cir. 1995) (unmodified term ―associating‖ not limited to explicit

21

association)). Here, there are no modifiers in the claim language suggesting that the video monitor

22

must be physically separate from the camera, nor is there any reason to believe that the language

23

should be understood that way. Indeed, ―video monitor‖ has no modifiers at all. Similarly, the

24

Federal Circuit ―has repeatedly and clearly held that it will not read unstated limitations into claim

25

language.‖ Rambus Inc. v. Infineon Techs. Ag, 318 F.3d 1081, 1088 (Fed. Cir. 2003) (citation

26

omitted). Nothing in the patent suggests a limitation of ―physically separate from the camera

27

body‖ for the monitor, and the Court will not presume one.

28

Although not dispositive, the patent prosecution history supports this construction.
7

1

Fujifilm cites to the original Claim 2 of the ‗427 Patent‘s earlier application, which uses the same

2

exact language as the disputed term. Br. 5; Cangro Decl. Ex. 3 at 28. In rejecting the proposed

3

claim based on prior art, the examiner referenced United States Patent No. 4,740,828, which

4

disclosed a ―monitor [that] is the built-in monitor of a video camera – which implies that it is used

5

as the camera viewfinder.‖ Cangro Decl. Ex. 4 at 2. Interpreting the same language as the

6

disputed term here, the patent examiner apparently understood that the ―viewfinder‖ could be a

7

built-in monitor. And in responding to the rejection, the applicant never argued that the

8

viewfinder did not meet the proposed claim because it was a built-in monitor, but only argued that

9

its monitor is a color monitor while the prior art disclosed a black and white monitor. Cangro
Decl. Ex. 6 at 14. Neither examiner nor applicant thought that the ―viewfinder‖ had to be

11
United States District Court
Northern District of California

10

physically separate. As Fujifilm points out, the applicant could have simply disputed the built-in

12

requirement but did not and instead objected to the examiner‘s rejection on a different issue.

13

Fujifilm‘s Reply (Dkt. No. 49) (―Reply‖) 3. The Court finds the applicant‘s silent acquiescence to

14

be persuasive evidence that supports construing the disputed term to allow for built-in monitors.
The ‗427 Patent‘s preferred embodiment also supports this construction. As the patent

15
16

says, ―The monitor 29 is connected to the reproduction 26,‖ and ―this connection 27 may have a

17

fixed configuration or a separable configuration.‖ ‗427 Patent 3:45-48. The fact that the

18

connection between the monitor and the rest of the camera can be ―fixed‖ (as opposed to

19

―separable‖) supports the conclusion that the patent contemplates built-in monitors as well as

20

physically separate ones. To construe the patent as requiring built-in monitors may exclude the

21

preferred embodiment, which the Federal Circuit has cautioned against. Vitronics, 90 F.3d at

22

1583.

23

Motorola disagrees with Fujifilm‘s understanding of the prosecution history. It argues that

24

an applicant need not ―invoke every conceivable reason a piece of prior art is not relevant.‖

25

Response 19. It asserts that because ―[t]he Applicant for the ‗427 patent, having distinguished [the

26

prior art] on two grounds already, apparently saw no need to delve into the Examiner‘s apparent

27

misunderstanding of the required monitor location,‖ though Motorola does not explain why those

28

two grounds should be credited but not this one. Motorola argues that the applicant ―simply
8

1

ignored an issue it had no obligation or need to address.‖ Response 20.
While Motorola correctly notes that the Federal Circuit has said that ―[a] patentee is not

2

required to fight tooth and nail every possibly adverse thought an examiner commits to paper,‖ the

4

court goes on to say that ―[w]hether the patentee chooses to dispute the examiner‘s view of matters

5

is relevant to claim interpretation, for there a court may need to ascertain exactly what subject

6

matter was actually examined and allowed by the PTO.‖ TorPharm, Inc. v. Ranbaxy Pharm., Inc.,

7

336 F.3d 1322, 1330 (Fed. Cir. 2003). A court may ―equate an inventor‘s acquiescence to the

8

examiner‘s narrow view of patentable subject matter with abandonment of the rest.‖4 Id. Those

9

principles apply here. The prosecution history supports construing the term according to its plain

10

and ordinary meaning, as Fujifilm proposes. Indeed, the very fact that the examiner was aware of

11
United States District Court
Northern District of California

3

built-in monitors suggests that a person of ordinary skill in the art would not assume that the

12

viewfinder must be physically separate from the camera.
Motorola quotes from the preferred embodiment‘s description as stating that ―the camera 1

13
14

may have an optical viewfinder thereon in addition to the independent viewfinder.‖ ‗427 Patent

15

3:48-50. Based on this, Motorola argues that the possibility of an ―optical view finder thereon‖ in

16

addition to an ―independent viewfinder‖ means that the monitor is remote from the camera. Br.

17

17. However, there is no support in the language of the patent or its prosecution history for the

18

understanding that an ―independent‖ viewfinder means that it is physically separate from the

19

camera, nor does the fact that the optical view finder is ―thereon‖ the camera change this.5 Even if

20

Motorola‘s understanding of ―optical viewfinder thereon in addition to the independent

21

viewfinder‖ is correct, its ultimate conclusion is belied by the fact that the preferred embodiment

22

only says that the camera ―may‖ have such a configuration, not ―must.‖ ‗427 Patent 3:49-50. 6

23
4

24
25
26
27
28

Motorola argues that in this case, the examiner was taking a broader view, not a ―narrower‖ one,
and thus estoppel should not apply. Response 19-20. While Motorola is correct that the Federal
Circuit used the word ―narrow,‖ the Court does not take such an unnecessarily constrained
understanding of the principle articulated in TorPharm.
5
The parties dispute whether the fact that among the patent‘s benefits are that the invention
―eliminates the need [for] a high-resolution monitor and . . . the space for the installation of a
monitor‖ suggests that the monitor is or is not physically separate. Br. 6; Response 17-18; ‗427
Patent 12:10-13. The Court does not find this issue dispositive either way because these benefits
are not inconsistent with either party‘s position.
6
The patent states that one ―object of the present invention‖ is ―to provide an electronic still
9

1
2
3
4

C. “imaging device for shooting a subject and outputting (a) corresponding (first
color image signal having a high resolution/high resolution pixel data) to be
recorded in a recording medium”
The third disputed term is found in Claims 1, 6, 14, and 15.
Fujifilm’s Construction
Not subject to § 112 ¶ 6, but if so ordered:

Motorola’s Construction
Subject to § 112 ¶ 6

Function: Capturing the image of a subject
and outputting a corresponding (high
resolution color image signal / high-resolution
pixel data) to be recorded in memory

Function: Shooting a subject and outputting
high resolution color image data

5
6
7
8
9

Structure: Imaging device 12: a high or low
resolution CCD; and equivalents thereof

Structure: Imaging device 12; and equivalents
thereof
The parties dispute whether the term is governed by § 112, ¶ 6.

11
United States District Court
Northern District of California

10

The Court finds that ―imaging device‖ is not subject to § 112, ¶ 6. ―The use of the term

12

‗means‘ triggers a rebuttable presumption that § 112, ¶ 6 governs the construction of the claim

13

term.‖ Inventio AG v. ThyssenKrupp Elevator Amers. Corp., 649 F.3d 1350, 1356 (Fed. Cir.

14

2011) (citation omitted). Where the term ―means‖ is not used, the court must presume that § 112,

15

¶ 6 does not apply. Id. Because the word ―means‖ has ―become closely associated with means-

16

plus-function claiming. . . . the presumption . . . is a strong one that is not readily overcome.‖ Id.

17

Indeed, ―the use of the word ‗means‘ is central to the analysis.‖ Id. (emphasis added). ―[T]he

18

presumption can be overcome if the challenger demonstrates that the claim term fails to recite

19

sufficiently definite structure or else recites function without reciting sufficient structure for

20

performing that function.‖ Id. (citations and quotation marks omitted). To determine whether the

21

patent discloses sufficient structure, a court may look to ―the words of the claims themselves, the

22

written description, the prosecution history, and any relevant extrinsic evidence.‖ Id. at 1356.

23

The Court finds that Motorola has not overcome the presumption that ―imaging device‖ is

24

not subject to § 112, ¶ 6. Motorola‘s argues that the term constitutes functional claiming. The

25

Court disagrees. The specification discloses a structure, imaging device 12, which is a ―solid-state

26
27
28

camera capable of being controlled at a remote place with a monitor serving as a viewfinder.‖
‗427 Patent at 1:57-59. Motorola argues that this is ―impossible‖ if the monitor was not external.
Response 18. On the contrary, having a camera with a built-in monitor is not inconsistent with
that objective.
10

1

imaging device‖ with a ―light-sensitive surface‖ that ―outputs an electric signal representative of

2

the incident image.‖ ‗427 Patent 2:57-61. Such a description is structural, not functional, and is

3

likely to be understood as such by a person of ordinary skill in the art. ―[T]his is not a case where

4

a claim nakedly recites a ‗device‘ and the written description fails to place clear structural

5

limitations on the ‗device.‘‖ Inventio, 649 F.3d at 1359. Motorola has provided no evidence or

6

persuasive argument to the contrary.

7

II.

8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18
19
20
21

22
23
24
25
26
27
28

PATENT NO. 7,327,886
This patent discloses a digital camera that can detect whether a face is included in the

frame of a moving image. The detection repeats until a picture is taken. The most recent location
of the face is recorded so that the image can be processed later, e.g., to correct red-eye effect.
A. “face judgment means for performing processing, in a predetermined time
interval, for judging whether a human face is included in a frame included in the
moving image until a positive result is obtained in the judgment”
The disputed term is found in Claim 1.
Fujifilm’s Construction
Function: performing processing, in a
predetermined time interval, for judging
whether a human face is included in a frame
included in the moving image until a positive
result is obtained in the judgment

Motorola’s Construction
Function: same

Structure: CPU 12 programmed to calculate a
first characteristic value C1, which is used for
discriminating the face from a single frame Fri,
and if found, calculating a second
characteristic value C2 from an image within a
facial candidate extracted from the frame Fri.
If it is judged that a facial candidate is not
included in the frame Fri in the first
discrimination or the second discrimination,
the CPU 12 judges that a face is not included
in the frame Fri and the CPU 12 performs the
first and second discrimination on a frame
Fri+1 after a predetermined time until a
positive result is obtained; and equivalents
thereof

Structure: CPU 12 programmed to perform the
steps of Fig. 14 of the ‗886 Patent, repeatedly
if necessary, until a positive result is
obtained as outlined below; and equivalents
thereof
1. Perform steps S31-S33.
Step S31 - calculate a first characteristic value
C1 as the directions of the gradient vectors in
the frame Fri, for each of a plurality of stages
of enlargement or reduction and rotation of the
frame Fri. This step requires preparatory steps
a and b:
a. perform filtering processing on a frame Fri
by using an edge detection filter in horizontal
and vertical directions; and
b. compute a gradient vector K at each pixel
11

1
2

based on the magnitude H of the edge in the
horizontal direction and the magnitude V of the
edge in the vertical direction at each pixel of
the frame Fri.

3
4
5
6
7
8
9
10

Step S32 - read first reference data R1 from
system memory, the first reference data R1
defining a discrimination condition for the
combination of the first characteristic value C1
at each pixel include in each of a plurality of
kinds of pixel groups including a plurality of
pixels, selected from a sample image.
Step S33 - use the first characteristic value C1
to perform a first discrimination for
discriminating a face from the frame Fri.

United States District Court
Northern District of California

11

2. If it is judged that a facial candidate is found
in the image Fri, perform steps S34-S38.

12

Step S34 - extract the facial candidate.

13
14
15
16
17
18
19
20
21

Step S35 - calculate a second characteristic
value C2 as the directions and magnitudes of
the gradient vectors within the facial candidate,
for each of a plurality of stages of enlargement
or reduction and rotation of the facial
candidate.
Step S36 - normalize the magnitude of the
gradient vector K of the second characteristic
value C2 by obtaining a histogram of the
magnitudes of the gradient vectors K at all
pixels in the facial candidate, and smoothing
the histogram so that the magnitudes of the
gradient vectors K are evenly distributed to all
the range of values.

22
23
24
25
26
27
28

Step S37 - read second reference data R2 from
system memory, the second reference data R2
defining a discrimination condition for the
combination of the second characteristic value
C2 at each pixel include in each of a plurality
of kinds of pixel groups including a plurality of
pixels, selected from a sample image.
Step S38 - use the second characteristic value
C2 to perform a second discrimination to
12

1
2
3
4
5
6

discriminate whether the facial candidate is a
face.
3. If the facial candidate is a face, perform step
S39 - judge that the frame includes a face.
4. If it is judged that the frame does not include
a facial candidate, or that a facial candidate is
included but is not a face, perform step S40 judge that the frame Fri does not include a
face.

7
8
9
10

United States District Court
Northern District of California

11
12
13

4. If it is judged that the frame does not include
a facial candidate, or that a facial candidate is
included but is not a face, perform step S40 judge that the frame Fri does not include a
face.
The parties agree that this is a means-plus-function claim. They also agree to the same
function, but disagree about the corresponding structure.
The Court adopts Motorola‘s construction and finds that Fujifilm‘s construction does not

14

adequately capture the disclosed structure. ―In a means-plus-function claim in which the disclosed

15

structure is a computer, or microprocessor, programmed to carry out an algorithm, the disclosed

16

structure is not the general purpose computer, but rather the special purpose computer

17

programmed to perform the disclosed algorithm.‖ WMS Gaming, Inc. v. Int’l Game Tech., 184

18

F.3d 1339, 1349 (Fed. Cir. 1999). Because a patentee does not need to disclose every component

19

of the structure, a patentee may ―express that procedural algorithm in any understandable terms

20

including as a mathematical formula, in prose, or as a flow chart, or in any other manner that

21

provides sufficient structure.‖ Typhoon Touch Techs., Inc. v. Dell, Inc., 659 F.3d 1376, 1385

22

(Fed. Cir. 2011) (citation and internal quotations omitted). The proper corresponding algorithm

23

consists only of the portion of the written description ―necessary to perform the claimed function.‖

24

Micro Chem., Inc. v. Great Plains Chem. Co., 194 F.3d 1250, 1258 (Fed. Cir. 1999); see also

25

Cardiac Pacemakers, Inc. v. St. Jude Med., Inc., 296 F.3d 1106, 1119 (Fed. Cir. 2002) (stating

26

that a ―corresponding structure need not include all things necessary to enable the claimed

27

invention to work. . . . [but it] must include all structure that actually performs the recited

28

function‖). ―[T]he patent need only disclose sufficient structure for a person of skill in the field to
13

1

provide an operative software program for the specified function.‖ Micro Chem., 194 F.3d at

2

1258.

3

The Federal Circuit has two lines of cases: one in which the specification discloses no

4

algorithm, and another in which the defendant contends that the disclosure is inadequate. Noah

5

Sys., Inc. v. Intuit Inc., 675 F.3d 1302, 1313 (Fed. Cir. 2012). Here, because both parties agree

6

that the patent discloses some structure, ―the question is whether the disclosed algorithm [as

7

construed by each party], from the viewpoint of a person of ordinary skill, is sufficient to define

8

the structure and make the bounds of the claim understandable.‖ Id. The structure ―must provide

9

―an adequate disclosure showing what is meant by th[e] claim language.‖ Id. at 1312 (brackets
omitted). A patentee is ―not required to produce a listing of source code or a highly detailed

11
United States District Court
Northern District of California

10

description of the algorithm to be used to achieve the claimed functions.‖ Aristocrat Techs.

12

Australia Pty Ltd. v. Int'l Game Tech., 521 F.3d 1328, 1338 (Fed. Cir. 2008). However, ―purely

13

functional language, which simply restates the function associated with the means-plus-function

14

limitation, is insufficient to provide the required corresponding structure.‖ Noah Sys., 675 F.3d at

15

1317. ―The inquiry is whether one of skill in the art would understand the specification itself to

16

disclose a structure, not simply whether that person would be capable of implementing a

17

structure.‖ Biomedino, LLC v. Waters Techs. Corp., 490 F.3d 946, 953 (Fed. Cir. 2007).

18
19

1. Fujifilm’s Proposed Structure Is Insufficient.
Fujifilm argues that its construction explains how CPU 12, which functions as the face

20

judgment means, performs the claimed function of ―judging whether a human face is included in a

21

frame included in the moving image.‖ Reply 17; ‗886 Patent at 7:20-37. When the camera is

22

taking a picture, the CPU 12 executes a program for judging whether a face is included in a single

23

frame of a moving image by calculating a ―first characteristic value C1, which is used for

24

discriminating the face, from a single frame Fri.‖ ‗886 Patent at 7:33-35. Next, it ―calculates a

25

second characteristic value C2 from an image within a facial candidate, extracted from the frame

26

Fri.‖ ‗886 Patent at 7:35-37. This process is repeated until a face is discriminated using the

27

values. If the CPU 12 determines that a facial candidate is not in the frame Fri in the first

28

discrimination, or even if it is, if the facial candidate is not in the second discrimination, then the
14

1

CPU 12 determines that a face is not included in the frame Fri. In that case, the CPU 12 performs

2

the first and second discrimination on a frame Fri+1 after some predetermined time to judge

3

whether a face is in the frame Fri+1 in a similar manner. ‗886 Patent at 14:31-44. If a face is

4

discriminated in both discriminations, then the CPU 12 judges that a face is included in the frame

5

Fri. Id.
Fujifilm‘s proposed construction does not sufficiently identify the disclosed structure. It is

6
7

mostly functional language that does not recite an algorithm that a person of ordinary skill in the

8

art can identify and use. As the Federal Circuit explained, ―In cases involving a computer-

9

implemented invention in which the inventor has invoked means-plus-function claiming, this court
has consistently required that the structure disclosed in the specification be more than simply a

11
United States District Court
Northern District of California

10

general purpose computer or microprocessor. . . . to avoid purely functional claiming.‖ Aristocrat

12

Techs., 521 F.3d at 1333. A proposed structure that ―simply references a computer that is

13

programmed so that it performs the function in question, which is to say that the function is

14

performed by a computer that is capable of performing the function,‖ is insufficient. Id. at 1334.

15

Here, the parties agree that the CPU‘s function is to ―judg[e] whether a human face is included in

16

a frame.‖ However, Fujifilm‘s construction essentially states that the CPU is ―programmed to

17

calculate a [] characteristic value [], which is used for discriminating the face.‖ This is nothing

18

more than simply restating the function of the computer without disclosing an algorithm for

19

performing that function. Especially telling is the fact that Fujifilm‘s construction says that the

20

CPU is ―programmed to calculate‖ its function, without adequately disclosing the structure or

21

method for performing such a calculation or such programming, effectively leaving the CPU a

22

general purpose computer. The construction does precisely what the algorithm requirement seeks

23

to prevent because it ―imposes no limitation whatever, as any general purpose computer must be

24

programmed.‖ Aristocrat Tech., 521 F.3d at 1334.
The Federal Circuit recently addressed whether an algorithm disclosed sufficient structure

25
26

in Ibormeith IP, LLC v. Mercedes-Benz USA, LLC, 732 F.3d 1376 (Fed. Cir. 2013).7 The patent

27
7

28

The Court recognizes that the Federal Circuit issued this opinion after the parties‘ briefing and
Markman hearing, though Motorola subsequently identified it as supplemental authority. Dkt. No.
15

1

there involved a ―sleepiness detection‖ device for vehicles and claimed ―computational means‖ for

2

calculating a driver‘s sleepiness, the corresponding structure of which was an algorithm. The

3

plaintiff argued that a table in the patent with various values and an ―equation‖ that added up those

4

values constituted the corresponding structure. The district court disagreed and held that, although

5

one skilled in the art might understand those items to be the corresponding structure, neither the

6

table nor any other clearly identified part of the specification ―states the steps to obtain the listed

7

elements‘ scores, or how to weight them according to their relative importance.‖ Id. at 1378-79

8

(quoting Ibormeith IP, LLC v. Mercedes-Benz USA, LLC, 889 F. Supp. 2d 677, 687-88 (D.N.J.

9

2012)) (emphasis added). Thus, the court found the ―algorithm‖ insufficient and the relevant

10

claims invalid.

United States District Court
Northern District of California

11

The Federal Circuit affirmed and concluded that ―the disclosed algorithm does not

12

adequately define the structure.‖ Id. The court stated that the table only provided ―what looks like

13

a mathematical equation‖ with ―no disclosure of even a single concrete relationship between the

14

various factors that are used to compute an outcome.‖ Id. at 1380-81. The table ―merely indicates

15

that weighting . . . occurs‖ but does not ―disclose how to perform the weighting,‖ and the Federal

16

Circuit credited the defense expert‘s statement that ―a person of ordinary skill in the art ‗would

17

need to devise his or her own method . . . based on the factors generally disclosed‘‖ in the patent.

18

Id. at 1381-82 (emphasis added). Thus, the Federal Circuit held, ―A description of an algorithm

19

that places no limitations on how values are calculated, combined, or weighted is insufficient to

20

make the bounds of the claim understandable.‖ Id. at 1382.

21

Fujifilm‘s construction exhibits the same defects as the alleged algorithm in Ibormeith.

22

Here, Fujifilm‘s proposed structure is ―programmed to calculate a [ ] characteristic value‖ and

23

―judge[s whether] a facial candidate is not included in the frame Fri,‖ but Fujifilm‘s construction

24

does not explain how to calculate the characteristic values or how discrimination occurs. As the

25

district court in Ibormeith recognized, an algorithm must ―state the steps to obtain‖ the values to

26

be calculated. Ibormeith IP, LLC v. Mercedes-Benz USA, LLC, 889 F. Supp. 2d 677, 687-88

27
28

64. However, the Court‘s conclusion would have been the same and was only reaffirmed by
Ibormeith.
16

1

(D.N.J. 2012). However, Fujifilm‘s construction ―merely indicates‖ that the calculation of

2

characteristic values ―occurs,‖ but does not ―disclose how to perform‖ the calculation. Ibormeith

3

IP, 732 F.3d at 1381. The same is true for judging whether a face is included, i.e., discrimination.

4

Indeed, Fujifilm‘s construction only provides ―what looks like a mathematical equation,‖ but

5

which, stripped of its jargon, says little that would guide a person of ordinary skill in the art, and

6

Fujifilm provides no evidence (e.g., by declaration) that it would. Id. at 1380. What Fujifilm

7

states in its Reply brief in support of its construction displays the central problem identified in

8

Ibormeith: ―Motorola complains that Fujifilm‘s construction does not explain how the first and

9

second characteristic value are calculated or how to perform the first and second discrimination.
The additional details of how the algorithm performs the claimed function, however, are not

11
United States District Court
Northern District of California

10

necessary to make it complete.‖ Reply 11 (original emphasis). On the contrary, as the Federal

12

Circuit held, ―A description of an algorithm that places no limitations on how values are calculated

13

. . . is insufficient to make the bounds of the claim understandable.‖ Id. at 1382. Fujifilm‘s

14

construction leaves the calculation of its values unbounded and is therefore insufficient.
Aristocrat Technologies Australia Pty Ltd. v. International Game Technology is also

15
16

instructive.8 521 F.3d at 1329. The patent in that case involved an electronic slot machine that

17

allows a player to select winning combinations of symbol positions. In assessing a means-plus-

18

function term, the trial court characterized the function as a means ―(1) to control images

19

displayed on the display means; (2) to pay a prize when a predetermined combination of symbols

20

matches the symbol positions selected by the player; and (3) to define the pay lines for the game

21

according to each possible combination of the selected symbol positions.‖ Id. at 1331. The

22

Federal Circuit affirmed the trial court‘s conclusion that the proffered structure—―any standard

23

microprocessor base [sic] gaming machine [with] appropriate programming‖—insufficiently

24

disclosed structure because it ―simply references a computer that is programmed so that it

25

performs the function in question, which is to say that the function is performed by a computer

26
8

27
28

The Federal Circuit ultimately found that no algorithm was disclosed at all in this case, contrary
to the plaintiff‘s assertion. Although the parties here do not dispute whether the ‗866 Patent
discloses some structure, the principles in Aristocrat Technologies are nonetheless instructive to
determine whether Fujifilm‘s construction of the structure is sufficient.
17

1

that is capable of performing the function.‖ Id. at 1334 (original additions). Further, the Federal

2

Circuit rejected the argument that the algorithm is that the system would ―pay a prize when a

3

predetermined combination of symbols is displayed in a predetermined arrangement of symbol

4

positions selected by a player‖ because ―that language simply describes the function to be

5

performed, not the algorithm by which it is performed.‖ Id. (emphasis added). The court also

6

found the proffered structure of ―defining a set of predetermined arrangements for a current game

7

comprising each possible combination of the symbol position selected by the player which have

8

one and only one symbol position in each column of the display means‖ to be wanting for the

9

same reason. Id. The plaintiff pointed the court to the patent‘s embodiments, tables, and figures
for an algorithm, but the court concluded that they were nothing more than ―pictorial and

11
United States District Court
Northern District of California

10

mathematical ways of describing the claimed function,‖ which ―is not enough to transform the

12

disclosure of a general-purpose microprocessor into the disclosure of sufficient structure to satisfy

13

section 112 paragraph 6.‖ Id. at 1335.

14

The structure offered by Fujifilm is similar to what the plaintiff in Aristocrat Technologies

15

offered. Fujifilm‘s construction does nothing more than repeat the function of the term and its

16

outcome. Here, where the agreed function is ―performing processing, in a predetermined time

17

interval, for judging whether a human face is included in a frame included in the moving image

18

until a positive result is obtained in the judgment,‖ it is insufficient to say that the structure

19

essentially (1) calculates whether a face is in a frame, (2) calculates again whether a face is in a

20

frame, and (3) repeats this process ―after a predetermined time until a positive result is obtained.‖

21

But this is what Fujifilm‘s construction proposes, and thus it fails. The Federal Circuit in Typhoon

22

Touch Technologies, Inc. v. Dell, Inc.—a case cited by Fujifilm—noted that its predecessor court

23

defined an algorithm as ―[a] fixed step-by-step procedure for accomplishing a given result.‖ 659

24

F.3d at 1385 (quoting In re Freeman, 573 F.2d 1237, 1245 (C.C.P.A. 1978)). Fujifilm has not

25

provided this.

26

Fujifilm argues that while other parts of the patent‘s specification includes examples of

27

how these values may be calculated, as well as descriptions of other functions performed by the

28

claimed inventions, these methods and ―specific discriminations are not required for the facial
18

1

judging to occur.‖ Br. 18. ―In particular, while calculation of C1 and C2 are necessary to perform

2

face judgment, the particular steps of generating the discriminators chosen from the various

3

possibilities are not requirements to perform the function.‖ Br. 18. Because additional specificity

4

from the embodiments is unnecessary to perform the claimed function, Fujifilm argues that they

5

should not be included in the structure corresponding to the face judgment means limitation. Br.

6

18. Fujifilm argues that its proposed construction ―closely tracks the language from the

7

specification.‖ Br. 18.

8

While Fujifilm is right in saying that it need only disclose an algorithm that performs the
necessary function, the problem with its ―algorithm‖ is that the ―algorithm,‖ in essence, only

10

repeats the function itself, albeit with variables. Fujifilm‘s construction does not define the

11
United States District Court
Northern District of California

9

structure and make the bounds of the claim understandable to a person of ordinary skill in the art.

12

Notably, Fujifilm does not argue that a person of ordinary skill in the art knows how to calculate a

13

value that discriminates a face from a frame, or how such discrimination occurs. As Ibormeith

14

teaches, it is not enough to identify ―an algorithm that places no limitations on how [its

15

component] values are calculated,‖ 732 F.3d at 1382, but Fujifilm does not explain how to derive

16

the characteristic values or how discrimination occurs. ―[W]hile it is true that the patentee need

17

not disclose details of structures well known in the art, the specification must nonetheless disclose

18

some structure.‖ Default Proof Credit Card Sys., Inc. v. Home Depot U.S.A., Inc. (d/b/a The

19

Home Depot), 412 F.3d 1291, 1302 (Fed. Cir. 2005) (citation omitted). Fujifilm could have

20

submitted a declaration or expert testimony that the elements recited in its construction are readily

21

known to a person of ordinary skill in the art, but it did not do so. It does not matter that

22

Fujifilm‘s construction closely tracks the patent‘s language if it tracks too little of it. There is

23

simply insufficient information to determine how the structure performs the claimed function. See

24

generally Function Media, L.L.C. v. Google, Inc., 708 F.3d 1310, 1319 (Fed. Cir. 2013).

25
26

2. The Court Adopts Motorola’s Construction.
After carefully considering the evidence cited by the parties, as well as the ‗886 Patent

27

itself, the Court adopts Motorola‘s construction. While long and seemingly unwieldy, the Court

28

finds that Motorola‘s construction accurately reflects the algorithm and structure necessary to
19

1

perform the function agreed upon by the parties. In particular, Motorola‘s construction explains

2

how the characteristic values are calculated and how the discrimination is performed, which

3

Fujifilm‘s construction does not do and which Ibormeith requires. As Fujifilm concedes, ―the

4

calculation of the characteristic values and the discriminations themselves . . . are necessary to

5

perform the recited function.‖ Br. 19 (emphasis added). Motorola‘s language carefully adheres to

6

the patent‘s language and does not import any unnecessary limitation.
Fujifilm argues that Motorola is incorporating most of the specification into the term ―face

8

judgment means‖ and is ―over-narrowing‖ the claim term. Reply 9. ―[S]pecific examples of how

9

[values and discriminations] are calculated or performed . . . are inappropriately included in

10

Motorola‘s proposed construction.‖ Br. 19. And because alternatives to reference data are

11
United States District Court
Northern District of California

7

disclosed, the inventor did not consider them necessary. Reply 10. Fujifilm points to the patent‘s

12

statement that ―it is obvious that other methods . . . may also be used.‖ 18:66-19:3. Even if the

13

―additional steps added by Motorola‘s construction could be characterized as being required for

14

the claimed invention to work,‖ Fujifilm argues, ― that would still be insufficient to include them

15

in the corresponding structure, which does not need to include all things necessary to enable the

16

claimed invention to work.‖ Br. 20 (citing Default Proof Credit Card Sys., Inc. v. Home Depot

17

U.S.A., Inc., 412 F.3d 1291, 1298 (Fed. Cir. 2005)).

18

Fujifilm‘s arguments are unpersuasive. Fujifilm is only partially correct when it says that

19

the patent discloses ―alternatives‖ (plural). Reply 10. With regard to Motorola‘s construction, the

20

only aspect of the algorithm for which the possibility of an alternative is disclosed and that

21

Fujifilm identifies is the method for obtaining reference data, not the use of reference data itself.

22

‗886 Patent 18:63-19:3. There, the patent discloses that ―it is obvious that other methods [for

23

obtaining reference data], such as a method for judging whether a shape, which conforms to a

24

facial template . . . may also be used‖ rather than the ―machine learning method‖ discussed in the

25

patent. Id. Thus, Fujifilm is wrong to say that ―[b]y describing alternatives to reference data, the

26

inventor did not consider reference data to be necessary for the algorithm.‖ Reply 10 (original

27
28
20

1

emphasis). The patent was only discussing one kind of reference data, not all reference data.9

2

Fujifilm provides no argument as to why reference data are unnecessary if the face detection

3

means is to discriminate the presence of a face in the frame. Other than this example, Fujifilm

4

identifies no other instance in which the patent discusses alternatives to the method disclosed. In

5

any event, Motorola‘s construction does not require any particular kind of reference data.

6

Similarly, Fujifilm is incorrect in criticizing Motorola‘s inclusion of ―preparatory steps‖ in

7

its construction. Reply 9. While Motorola‘s use of those words may lead one to believe that such

8

steps are unnecessary to perform the function, reading the patent reveals that the steps discussed

9

are in fact part of the calculation of the characteristic values, and therefore necessary. The
calculation of the characteristic values depends upon the directions of gradient vectors, and the

11
United States District Court
Northern District of California

10

―preparatory steps‖ explain how those vectors are calculated. ‗866 Patent 7:33-54. This comports

12

with Ibormeith‘s requirement that the calculation of values must somehow be limited. 732 F.3d at

13

1382. As above, Fujifilm provides no argument why these are not necessary for the algorithm.

14

Fujifilm rightly quotes the Federal Circuit as stating that ―corresponding structure need not include

15

all things necessary to enable the claimed invention to work.‖ Reply 20 (citing Default Proof

16

Credit Card Sys., 412 F.3d 1291 at 1298). However, the Federal Circuit‘s full sentence says,

17

―While corresponding structure need not include all things necessary to enable the claimed

18

invention to work, it must include all structure that actually performs the recited function.‖

19

Motorola‘s construction includes all structure that actually performs the recited function while

20

Fujifilm‘s does not.

21

B. “face detection means for detecting a facial position in a frame, which is judged to
include a face, if the face judgment means judges that the face is included in the
frame”

22
23
24
25
26
27
28

The disputed term is found in Claim 1.
Fujifilm’s Construction
Function: detecting a facial position in a
frame, which is judged to include a face, if the
face judgment means judges that the face is
included in the frame
9

Motorola’s Construction
Function: detecting a facial position in a
frame, if it has been previously determined by
the face judgment means that a face is included
in the frame

Fujifilm appears to recognize this on pages 18-19 of its opening brief.
21

1
2
3

Structure: CPU 12 programmed to obtain the
coordinate values at four corners of the
rectangle enclosing the face; and equivalents
thereof

4
5
6
7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18
19
20
21

22
23
24
25
26
27

Structure: (1) CPU 12 programmed to obtain
the coordinate values at four corners of the
region of 30x30 pixels, which correspond to
the position of the mask M or (2) CPU 12
programmed to obtain the center position,
which is the coordinate values of the
intersection of the diagonal lines of the mask
M and the length of a radius of a circle with its
center at the center position; and equivalents
thereof

The parties agree that ―face detection means‖ is a means-plus-function claim, but disagree
about the structure and function.
Function
The parties disagree about whether the system must detect a face before detecting its
location.
The Court adopts Motorola‘s construction. Motorola‘s construction, which contains the
predicate that ―it has been previously determined by the face judgment means that a face is
included in the frame,‖ because it is supported by the claim language. As the term states that
detection of the facial position (or location) in a frame occurs in a frame ―which is judged to
include a face,‖ the language suggests that the ―judging,‖ i.e., detecting the presence of a face, has
already occurred. Thus, it was ―previously determined by the face judgment means that a face is
included in the frame.‖ The conditional ―if‖ and the past tense in ―judged‖ and ―included‖ in the
third clause of the disputed term—much like the past tense in the second clause—support the
understanding that the face-detection occurs before the location-detection. In addition, when
Claim 1 discusses the ―control means‖ and references the face judgment means, it states that
―judgment is made as to whether the face is included . . . , and if the judgment is positive, the
facial position is detected.‖ ‗886 Patent 19:29-32. The claim language therefore shows that
detecting a face occurs before detecting its location.
Other portions of the patent‘s specification clarify that detecting a face occurs before
detecting a facial location. To begin, Figure 12 separates judging whether a face is present from
detecting the face‘s location, placing the former earlier in time than the latter, thus suggesting that

28
22

1

one precedes the other. Additionally, the abstract10 states, ―If it is judged that a face is included in

2

a frame, the facial position is detected in the frame, and stored. Then, judgment is made as to

3

whether a face is included in the next frame after predetermined time. If the judgment is positive,

4

the facial position is detected.‖ This process is repeated until a picture is taken. Finally, the

5

Summary of the Invention also reflects that detecting a face occurs before detecting its location.

6

‗886 Patent 2:13-16.
The preferred embodiment supports Motorola‘s construction. It states, ―When it is judged

7
8

that the face is included in the frame FRI . . . at which it was discriminated that the face was

9

included . . . the facial position is detected . . . .‖ ‗886 Patent 14:44-50 (emphasis added). Further,
―If it is judged that a face is not included . . . processing goes back to step S12. [But i]f it is

11
United States District Court
Northern District of California

10

judged that a face is included, the facial position is detected . . . .‖ ‗886 Patent 16:59-63 (emphasis

12

added). Additionally, the preferred embodiment discloses, ―If it is judged that a face is included,

13

the processing goes back to step S14, and the facial position is detected.‖ ‗886 Patent 17:22-23

14

(emphasis added). Taken as a whole, the patent‘s language strongly suggests that the inventor

15

intended that detecting a face occur before detecting its location, and that a person of ordinary skill

16

in the art would understand the term that way.

17

Structure

18

The parties agree that the structure includes a general purpose computer, CPU 12, but

19

disagree about the algorithm performed by it. Motorola‘s proposed construction includes an

20

alternative method of detecting a facial position, and Fujifilm argues that the alternative does not

21

comport with the specification. However, Fujifilm does not object to inclusion of this alternative

22

as long as it is clear that it is only an alternative and tracks the exact language of the specification.

23

Br. 22; Reply 12.
Based on this, the Court construes the structure as follows: ―CPU 12 programmed to

24
25

obtain the coordinate values at four corners of the rectangle enclosing the face; or CPU 12

26
27
28

10

The abstract may be considered in claim construction. See MEMS Tech. Berhad v. Int’l Trade
Comm’n, 447 F. App‘x 142, 151 (Fed. Cir. 2011); Netcraft Corp. v. eBay, Inc., 549 F.3d 1394,
1399 (Fed. Cir. 2008).
23

1

programmed to obtain the coordinate values at four corners of the region of 30x30 pixels, which

2

correspond to the position of the mask M; or CPU 12 programmed to obtain the center position,

3

which is the coordinate values of the intersection of the diagonal lines of the mask M and the

4

length of a radius of a circle with its center at the center position; and equivalents thereof.‖

5

C. “detecting a facial position in a frame, which is judged to include a face, if it is
judged that the face is included in the frame”

6

The disputed term is found in Claim 11.

7
8
9
10

United States District Court
Northern District of California

11

Fujifilm’s Construction
if the frame includes a face, detect the location
of the face within the frame

Motorola’s Construction
identifying the location of a human face within
frame if it has been previously determined that
the frame contains a human face

[for claim 1, this term should not be construed
separately from the ―face detection means‖
element]

12

The parties disagree about whether there is an ambiguity in the claim language: Motorola

13

argues that there is, Fujifilm argues that there is not. Br. 23; Response 11. Fujifilm argues that

14

Motorola‘s proposed construction attaches the requirement that ―it has been previously determined

15

that the frame contains a human face,‖ which ―adds ambiguity as to who or what is doing the

16

determination, and ―when this determination is to be performed and to what the performance is

17

prior.‖ Br. 23. Motorola argues that the determination is done ―earlier in time‖ when ―judging

18

whether a human face is included in a frame.‖ Response 11.

19

The Court adopts Motorola‘s construction for the same reasons it adopts Motorola‘s

20

construction for ―face detection means‖ above. In addition, the inclusion of the modifier ―human‖

21

to ―face‖ is warranted by the corresponding limitation in Claim 11. ‗886 Patent 21:20.

22

III.

23

PATENT NO. 6,144,763
This patent describes a digital camera that captures an image, converts the signal to digital

24

values, which are then converted to an image format with luminance (brightness) and chrominance

25

(color) components, and stored in memory. The chrominance components are set to fixed values,

26

then luminance and chrominance values are compressed based on a standard format, e.g., JPEG,

27

and stored to a recording medium. If the camera is set to monochrome mode, the chrominance

28
24

1

data are altered, resulting in a monochrome image.
The disputed terms are found in Claims 1, 12, and 19.

2
3
4
5
6
7
8

Term
writing processed image
data output from a signal
processing circuit in a
storage, and reading the
processed image data out
of said storage in
preselected blocks
component by component
(claim 12)

Fujifilm’s Constructions
writing processed image data
that is outputted from a signal
processing circuit into a
storage, and reading that
processed image data from the
storage in preselected blocks on
a component-by-component
basis

Motorola’s Constructions
writing at least one frame of
processed image data output
from a signal processing circuit
in a storage, and before reading
the processed image data out of
said storage in preselected
blocks on a component-bycomponent basis

storage for storing the
processed image data and
allowing the processed
image data to be read out
in preselected blocks
component by component
(claims 1 and 19)

storage for storing the
processed image data and for
allowing the processed image
data to be retrieved from the
storage in preselected blocks
on a component-by-component
basis

storage holding at least one
frame of processed image data,
and for allowing the processed
image data to be retrieved from
the storage in preselected blocks
on a component-by-component
basis

9
10

United States District Court
Northern District of California

11
12
13

The parties disagree about whether there must be ―at least one frame‖ of ―processed image

14
15
16

data.‖
The Court adopts Fujifilm‘s proposed constructions. Contrary to Motorola‘s constructions,

17

there is nothing in the claim language to suggest that the ―processed image data‖ must constitute

18

―at least one frame.‖ The claims do not even use the word ―frame‖ at all. Rather, they only speak

19

of data. For example, Claim 1 references ―outputting resulting coded image data,‖ ―a signal

20

processing circuit for processing the image data to output processed image data,‖ and—as the

21

disputed portion states, ―storage for storing the processed image data and allowing the processed

22

image data to be read out in preselected blocks component by component.‖ ‗763 Patent 13:51-57.

23

Similarly, Claim 12 discloses ―[a] method of compressing and coding image data representative of

24

a color image,‖ which is done by ―writing processed image data output from a signal processing

25

circuit in a storage, and reading the processed image data out of said storage in preselected blocks

26

component by component.‖ ‗763 Patent 14:60-67. While the claims speak of ―data,‖ ―blocks,‖

27

and ―components,‖ they never mention frames.

28

On the other hand, the preferred embodiment does discuss frames—indeed, it speaks of
25

1

single frames repeatedly, indicating that the inventor could have used the term ―frame‖ in the

2

claims if he wished. As the Federal Circuit has stated, ―we have repeatedly warned against

3

confining [] claims to those embodiments‖ discussed in the patent. Phillips, 415 F.3d at 1323.

4

The Court does not do so here.
Motorola argues that ―[t]he specification describes capturing and processing an image of

5

data on a frame-by-frame basis.‖ Motorola argues that ―at least one frame of data is required

7

because anything less than a complete frame of data would only represent a portion (or fraction) of

8

the image data captured by the image sensor.‖ Response 23. It may very well be the case that

9

―anything less than a complete frame of data‖ would be a partial image. However, the Court is

10

confined by what the patent‘s claims disclose, as supported by the specification and prosecution

11
United States District Court
Northern District of California

6

history, and Motorola has pointed to no evidence except the preferred embodiment to support its

12

construction.

13

IV.

14

PATENT NO. 6,915,119
This patent describes a wireless telephone that communicates both over a cellular network

15

to a provider, e.g., Verizon, and through another wireless channel without such a provider, e.g.,

16

Wi-Fi or Bluetooth. This allows the telephone to receive data from a cellular link and then

17

transmit it over the provider-less channel.

18

The disputed term is “menu (comprises/includes) selections for [a], . . . , [y], and [z].”

19

The disputed term is found in Claims 1, 7, 12, 19, 32, and 38.

20

Fujifilm’s Construction
Plain and ordinary meaning;

21

22
23

Motorola’s Construction
A single screen displaying selections for each
of [a], . . . , [y], and [z]

Alternatively: menu includes options
representing the following: [a], . . . , [y], and
[z]

24

The parties disagree about whether a ―menu‖ must be on a single screen.

25

The Court finds that this term does not need construction. A juror would understand the

26

term based on its plain and ordinary meaning. There is nothing in the claim language to suggest

27

that a ―menu‖ must be on a ―single screen,‖ as Motorola argues. And Motorola has provided no

28

argument or evidence that a person of ordinary skill in the art would understand the term ―menu‖
26

1

to mean that. This appears to be a situation in which ―the ordinary meaning of claim language as

2

understood by a person of skill in the art may be readily apparent even to lay judges, and claim

3

construction in such [a] case[ would] involve[] little more than the application of the widely

4

accepted meaning of commonly understood words.‖ Phillips, 415 F.3d at 1314. More

5

importantly, it is not a situation in which the Court may depart from the plain meaning of the term

6

because the patent has its own definition of ―menu‖ or because of a disavowal during patent

7

prosecution. Thorner, 669 F.3d at 1365. Motorola is attempting to limit the scope of the claim,

8

but has not given a sufficient basis for doing so.
Motorola argues that the term ―‗menu‖ has a more limited meaning tha[n] just merely

9

displaying a selection [of] options to a user.‖ Response 24. It says that a menu‖ is not ―merely

11
United States District Court
Northern District of California

10

the display of a selection of options to a user that may appear either on one screen, or any number

12

of screens, irrespective of whether there is any relationship between the screens.‖ Id. But

13

Motorola does not explain why a menu must only be on one screen or why another selection of

14

items on another screen which is related to a first selection of items cannot all be ―a menu.‖
Motorola points to Figures 1 and 4 of the patent, noting that the images on the screen

15
16

display a menu, and thus argues that a menu must be on one page because its construction ―is

17

consistent with both the only menus depicted and disclosed in the specification.‖ Response 25.

18

That argument is unconvincing. Motorola does not explain why any of the selections on the

19

screen could not be submenus of the same menu. More importantly, Motorola‘s seeming attempt

20

to limit the scope of the claims based on particular embodiments violates well-established canons

21

of claim construction.
Motorola argues that Fujifilm‘s construction ―would render the term ‗menu‘ meaningless.‖

22
23

Response 25. The Court is unclear why a list of selections or options, as Fujifilm construes the

24

disputed term, is meaningless. Motorola also argues that the fact that the word ―menu‖ is

25

preceded by the terms ―a‖ or ―said‖ somehow supports its construction. Response 25. But

26

Motorola does not adequately explain why this necessarily means that a menu must be on one

27

screen.

28
27

1

V.

This patent shares the same specification as its parent, the ‗886 Patent. Br. 23; Response

2
3
4

PATENT NO. 8,306,285

11.
The disputed term is “face judgment device for performing processing, at

5

predetermined or varying time intervals, for judging whether a human face is included in a

6

frame of the moving image”. The disputed term is found in Claims 1 and 13.

7

Fujifilm’s Construction
Not subject to § 112 ¶ 6, but, if so ordered:

Motorola’s Construction
Subject to § 112 ¶ 6

Function: ―performing processing, at
predetermined or varying time intervals, for
judging whether a human face is included in a
frame of the moving image‖

Function: ―performing processing, at
predetermined or varying time intervals, for
judging whether a human face is included in a
frame‖

Structure: same as ―face judgment means‖ of
the ‗886 patent, above

Structure: same as ―face judgment means‖ of
the ‗886 patent, above

8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18

The parties dispute whether the term is governed by § 112, ¶ 6.
The Court finds that ―face judgment device‖ is not subject to § 112, ¶ 6. Given the
―strong‖ presumption that claim language lacking the word ―means‖ is not subject to § 112, ¶ 6,
the Court must determine whether the patent is so devoid of structure that the presumption is
overcome. Inventio AG, 649 F.3d at 1356 (citation omitted).
The claim language and written description disclose sufficient structure such that the

19

presumption remains that ―face judgment device‖ is not subject to § 112, ¶ 6. Here, the claim

20

language is ―face judgment device for performing processing, at predetermined or varying time

21

intervals, for judging whether a human face is included in a frame of the moving image.‖ Looking

22

to the written description, the patent discloses ―[o]peration programs of the CPU 12[] includ[e]

23

programs for performing various kinds of processing,‖ such as ―processing for judging whether a

24

face is included in an image represented by image data.‖ ‗285 Patent 6:47-51. In addition, ―the

25
26
27

CPU 12 reads out the programs for executing processing for judging whether a face is included in
the image . . . and executes the program. Accordingly, the CPU 12 functions as . . . a face
judgment means . . . .‖ ‗285 Patent 7:11-16. The patent proceeds to describe at length how the

28
28

1

CPU 12 conducts its processing. The claim language thus points one to the corresponding

2

structure in the written description. Motorola has not met its burden of overcoming the

3

presumption.
Motorola argues that several cases hold that ―device‖ is insufficient to connote structure:

4

Mass. Inst. of Tech. & Elecs. for Imaging, Inc. v. Abacus Software, 462 F.3d 1344 (Fed. Cir. 2006)

6

(―MIT‖); Vistan Corp. v. Fadei USA, Inc., No. 10-cv-4862-JCS, 2012 WL 1496099 (N.D. Cal.

7

Apr. 27, 2012); and Widevine Techs., Inc. v. Verimatrix, Inc., No. 07-cv-321, 2009 WL 3734106

8

(E.D. Tex. Nov. 4, 2009). In MIT, the Federal Circuit stated that terms like ―device‖ ―typically do

9

not connote sufficiently definite structure.‖ MIT, 462 F.3d at 1354. The court also found that the

10

modifier ―colorant selection‖ did not connote sufficient structure. Id. Similarly, Vistan held that

11
United States District Court
Northern District of California

5

―assembly‖ does not connote structure. Vistan, 2012 WL 1496099, at *15. The court in Vistan

12

also found that adding the modifier of ―active‖ to ―assembly‖ does not help it. Id. at *16. Finally,

13

in Widevine, the court found the terms ―first device‖ and ―second device‖ to be means-plus-

14

function limitations because they did not identify any structure. Widevine, 2009 WL 3734106, at

15

*14.

16

Motorola has still not rebutted the ―strong presumption‖ that the lack of the word ―means‖

17

in the claim renders ―face judgment device‖ a means-plus-function limitation. In Inventio, the

18

Federal Circuit held that ―[t]he term ‗modernizing device‘ presumptively connotes sufficiently

19

definite structure to those of skill in the art, and that presumption is a strong one that is not readily

20

overcome.‖ 649 F.3d at 1358. It stated that the term ―device‖ should not be construed to be a

21

means-plus-function limitation when the claims ―delineate the components that the [] device is

22

connected to, describe how the [] device interacts with those components, and describe the

23

processing that the [] device performs,‖ and when ―[t]he written descriptions additionally show

24

that the [] device conveys structure to skilled artisans‖ and does not ―nakedly recite[] a ‗device‘‖

25

or ―fail[] to place clear structural limitations on the ‗device.‘‖ Id. at 1359. The ‗285 Patent meets

26

all of these requirements set out in Inventio. Significantly, the Federal Circuit deemed as dicta its

27

comment in MIT that ―device‖ was a generic term that does not connote sufficient structure and

28

rejected applying that principle in Inventio. Id. at 1359. Inventio supports Fujifilm‘s construction.
29

To further the point, in construing another term alleged to be means-plus-function, the

1
2

Inventio court found that ―written descriptions refer[ring] to the computing unit as a computer,

3

where one of its functions is to store and execute a computer program product,‖ and ―explain[ing]

4

the steps that the computer program product performs‖ is sufficient to preserve the presumption.

5

Id. at 1359-60. This language is even more on point regarding the ‗285 Patent. Unlike the

6

extremely ambiguous terms in Vistan and Widevine, the term here, ―face judgment device,‖

7

coupled with the corresponding structure, are sufficient for the Court to conclude that § 112, ¶ 6

8

does not apply.
CONCLUSION

9

As stated above, the Court construes the disputed terms as follows:

10

United States District Court
Northern District of California

11
12

I.

PATENT NO. 5,734,427
1. “Thinning said first image signal to thereby produce a second image signal having a

13

low resolution” is ―reducing the resolution of the first image signal to create a low

14

resolution image signal representative of the subject.‖

15

2. “Thinning said high-resolution pixel data output from said imaging device to thereby

16

produce low-resolution image data” is ―reducing the resolution of the pixel data output

17

from the imaging device to create low-resolution image data representative of the subject.‖

18

3. “Thinning said pixel data” is ―reducing the resolution of the pixel data.‖

19

4. “Thinning said high-resolution image data output from said imaging device to

20

thereby produce low-resolution image data” is ―reducing the resolution of the image

21

data output from the imaging device to create low resolution image data representative of

22

the subject.‖

23

5. “Thinning said first image signal . . . to thereby produce a second image signal having

24

low resolution” is ―reducing the resolution of the image signal to create a low resolution

25

image signal representative of the subject.‖

26

6. “A viewfinder . . . implemented by a video monitor” needs no construction.

27
28
30

1

7. “Imaging device for shooting a subject and outputting (a) corresponding (first color

2

image signal having a high resolution/high resolution pixel data) to be recorded in a

3

recording medium” is not a means-plus-function limitation.

4

II.

PATENT NO. 7,327,886
1. “Face judgment means for performing processing, in a predetermined time interval,

6

for judging whether a human face is included in a frame included in the moving

7

image until a positive result is obtained in the judgment” is a means-plus-function

8

limitation, in which the function is ―performing processing, in a predetermined time

9

interval, for judging whether a human face is included in a frame included in the moving

10

image until a positive result is obtained in the judgment,‖ and the corresponding structure

11
United States District Court
Northern District of California

5

is:

12

―CPU 12 programmed to perform the steps of Fig. 14 of the ‗886 Patent, repeatedly

13

if necessary, until a positive result is obtained as outlined below; and equivalents

14

thereof.

15

1. Perform steps S31-S33.

16

Step S31 - calculate a first characteristic value C1 as the directions of the

17

gradient vectors in the frame Fri, for each of a plurality of stages of

18

enlargement or reduction and rotation of the frame Fri. This step requires

19

preparatory steps a and b:

20

a. perform filtering processing on a frame Fri by using an edge

21

detection filter in horizontal and vertical directions; and

22

b. compute a gradient vector K at each pixel based on the magnitude

23

H of the edge in the horizontal direction and the magnitude V of the

24

edge in the vertical direction at each pixel of the frame Fri.

25

Step S32 - read first reference data R1 from system memory, the first

26

reference data R1 defining a discrimination condition for the combination

27

of the first characteristic value C1 at each pixel include in each of a plurality

28
31

1

of kinds of pixel groups including a plurality of pixels, selected from a

2

sample image.

3

Step S33 - use the first characteristic value C1 to perform a first

4

discrimination for discriminating a face from the frame Fri.

5

2. If it is judged that a facial candidate is found in the image Fri, perform steps S34-

6

S38.

7

Step S34 - extract the facial candidate.

8

Step S35 - calculate a second characteristic value C2 as the directions and

9

magnitudes of the gradient vectors within the facial candidate, for each of a
plurality of stages of enlargement or reduction and rotation of the facial

11
United States District Court
Northern District of California

10

candidate.

12

Step S36 - normalize the magnitude of the gradient vector K of the second

13

characteristic value C2 by obtaining a histogram of the magnitudes of the

14

gradient vectors K at all pixels in the facial candidate, and smoothing the

15

histogram so that the magnitudes of the gradient vectors K are evenly

16

distributed to all the range of values.

17

Step S37 - read second reference data R2 from system memory, the second

18

reference data R2 defining a discrimination condition for the combination

19

of the second characteristic value C2 at each pixel include in each of a

20

plurality of kinds of pixel groups including a plurality of pixels, selected

21

from a sample image.

22

Step S38 - use the second characteristic value C2 to perform a second

23

discrimination to discriminate whether the facial candidate is a face.

24

3. If the facial candidate is a face, perform step S39 - judge that the frame includes

25

a face.

26

4. If it is judged that the frame does not include a facial candidate, or that a facial

27

candidate is included but is not a face, perform step S40 - judge that the frame Fri

28

does not include a face.‖
32

2. “Face detection means for detecting a facial position in a frame, which is judged to

2

include a face, if the face judgment means judges that the face is included in the

3

frame” is a means-plus-function limitation, in which the function is ―detecting a facial

4

position in a frame, if it has been previously determined by the face judgment means that a

5

face is included in the frame,‖ and the corresponding structure is ―CPU 12 programmed to

6

obtain the coordinate values at four corners of the rectangle enclosing the face; or CPU 12

7

programmed to obtain the coordinate values at four corners of the region of 30x30 pixels,

8

which correspond to the position of the mask M; or CPU 12 programmed to obtain the

9

center position, which is the coordinate values of the intersection of the diagonal lines of

10

the mask M and the length of a radius of a circle with its center at the center position; and

11
United States District Court
Northern District of California

1

equivalents thereof.‖

12

3. “Detecting a facial position in a frame, which is judged to include a face, if it is judged

13

that the face is included in the frame” is ―identifying the location of a human face within

14

frame if it has been previously determined that the frame contains a human face.‖

15

III.

PATENT NO. 6,144,763
1. “Writing processed image data output from a signal processing circuit in a storage,

16
17

and reading the processed image data out of said storage in preselected blocks

18

component by component” is ―writing processed image data that is outputted from a

19

signal processing circuit into a storage, and reading that processed image data from the

20

storage in preselected blocks on a component-by-component basis.‖
2. “Storage for storing the processed image data and allowing the processed image data

21

22

to be read out in preselected blocks component by component” is ―storage for storing

23

the processed image data and for allowing the processed image data to be retrieved from

24

the storage in preselected blocks on a component-by-component basis.‖

25

IV.

“Menu (comprises/includes) selections for [a], . . . , [y], and [z]” needs no construction.

26
27
28

PATENT NO. 6,915,119

V.

PATENT NO. 8,306,285
“Face judgment device for performing processing, at predetermined or varying time
33

1

intervals, for judging whether a human face is included in a frame of the moving

2

image” is not a means-plus-function limitation.
***

3
4

The Court will hold a status conference on Tuesday, December 17, 2013, at 2 p.m., in

5

Courtroom 2, 17th Floor, 450 Golden Gate Avenue, San Francisco, California, at which the parties

6

should be prepared to discuss case scheduling and any other matters that may require the Court‘s

7

attention.

8
9
10

United States District Court
Northern District of California

11

IT IS SO ORDERED.
Dated: November 18, 2013
______________________________________
WILLIAM H. ORRICK
United States District Judge

12
13
14
15
16
17
18
19
20
21

22
23
24
25
26
27
28
34