Ex Parte Moroney

Patent Trial and Appeal Board

Jun 7, 2013

11259597 (P.T.A.B. Jun. 7, 2013)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________________

Ex parte NATHAN M. MORONEY1
____________________

Appeal 2010-011109
Application 11/259,597
Technology Center 2600
____________________

Before KALYAN K. DESHPANDE, BRYAN F. MOORE, and
LARRY J. HUME, Administrative Patent Judges.

HUME, Administrative Patent Judge.

DECISION ON APPEAL
This is a decision on appeal under 35 U.S.C. § 134(a) of the Final
Rejection of claims 1-5, 7-11, 13-17, and 19-24. Appellant has previously
canceled claims 6, 12, and 18. We have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM.

1 The Real Party in Interest is Hewlett-Packard Development Company, LP.
(App. Br. 2.)
Appeal 2010-011109
Application 11/259,597

2
STATEMENT OF THE CASE 2
The Invention
Appellant’s invention relates to the naming of visual attributes such as
color, shape, and texture, and more particularly to an adaptive lexical
classification system. Spec. p. 1, ll. 6-9 (“Field of the Present Invention”).
Exemplary Claims
Claims 1 and 7 are exemplary claims representing aspects of the
invention which are reproduced below (emphasis added):
1. A method for assigning a lexical classifier to
characterize a visual attribute corresponding to an image
element forming part of an image, the method comprising the
following steps:
capturing an initial attribute value for the image element
with an electronic sensor;
scaling a database including a set of lexical classifiers
corresponding to a particular type of visual attribute to a
smaller subset of said database based on an intended
application, said database comprising thousands of lexical
classifiers given by human subjects to different visual
attribute values for image elements;
transforming the initial attribute value to a designated
lexical classifier by reference to said smaller subset; and

2 Our decision refers to Appellants’ Appeal Brief (“App. Br.,” filed
Mar. 15, 2010); Reply Brief (“Reply Br.,” filed Aug. 5, 2010); Examiner’s
Answer (“Ans.,” mailed June 11, 2010); Final Office Action (“FOA,”
mailed Nov. 16, 2009); and the original Specification (“Spec.,” filed
Oct. 25, 2005).
Appeal 2010-011109
Application 11/259,597

3
outputting the designated lexical classifier assigned to the
visual attribute with an output device.
7. A system for assigning a lexical classifier to
characterize a visual attribute corresponding to an image
element forming part of an image, the system comprising:
an input device for capturing an initial attribute value for
the image element;
a database providing a set of lexical classifiers
corresponding to a particular type of visual attribute;
a processor for applying a machine learning algorithm to
transform the initial attribute value to a designated lexical
classifier by generating a histogram for each lexical classifier
in a specified vocabulary based on usage of that lexical
classifier in said database; and
generating a set of fuzzy membership functions based
on said histograms; wherein said designated lexical classifier
is produced by said processor in accordance with said fuzzy
membership functions; and
an output device for communicating the designated
lexical classifier for subsequent applications.
Prior Art
The Examiner relies upon the following prior art in rejecting the
claims on appeal:
Beretta US 5,416,890 May 16, 1995
Prabhakar US 2002/0031268 A1 Mar. 14, 2002
Mojsilovic US 2006/0087517 A1 Apr. 27, 2006
Appeal 2010-011109
Application 11/259,597

4
Rejections on Appeal
1. Claims 1-5, 19-21, and 24 stand rejected under 35 U.S.C.
§ 103(a) as being unpatentable over the combination of Mojsilovic and
Beretta. Ans. 4.
2. Claims 7-8, 10-11, 13-14, and 16-17 stand rejected under 35
U.S.C. § 103(a) as being unpatentable over the combination of Mojsilovic
and Prabhakar. Ans. 7.
3. Claims 22-23 stand rejected under 35 U.S.C. § 103(a) as being
unpatentable over the combination of Mojsilovic, Beretta, and Prabhakar.
Ans. 11.
4. Claims 9 and 15 stand rejected under 35 U.S.C. § 103(a) as
being unpatentable over the combination of Mojsilovic, Prabhakar, and
Beretta. Ans. 12.
ISSUES AND ANALYSIS
We have reviewed the Examiner’s rejections in light of Appellant’s
arguments that the Examiner has erred. We disagree with Appellant’s
conclusions with respect to claims 1-5, 7-11, 13-17, and 19-24, and we adopt
as our own (1) the findings and reasons set forth by the Examiner in the
action from which this appeal is taken and (2) the reasons and rebuttals set
forth by the Examiner in the Examiner’s Answer in response to Appellant’s
Arguments. However, we highlight and address specific findings and
arguments regarding claims 1 and 7 for emphasis as follows.
Appeal 2010-011109
Application 11/259,597

5
1. Rejection of Claims 1-5 and 19-24
Issue 1
Appellant argues (App. Br. 10-13; Reply Br. 4-8) that the Examiner’s
rejection of claim 1 under 35 U.S.C. § 103(a) as being unpatentable over the
combination of Mojsilovic and Beretta is in error. These contentions present
us with the following issues:
Did the Examiner err in finding that Appellant’s claimed method is
unpatentable because:
(a) the combination of Mojsilovic and Beretta teaches or suggests the
step of “scaling a database including a set of lexical classifiers
corresponding to a particular type of visual attribute to a smaller subset of
said database based on an intended application, said database comprising
thousands of lexical classifiers given by human subjects to different visual
attribute values for image elements,” as recited in claim 1; and because
(b) by using databases, Mojsilovic teaches away from the claimed
database produced by collecting lexical classifiers from actual human
subjects?
Analysis
Issue 1(a): Appellant contends that “the prior art to Mojsilovic and
Beretta does not teach or suggest scaling a database comprising thousands of
lexical classifiers to a smaller subset of that database as claimed.” App.
Br. 11. Appellant further contends that “the cited prior art merely teaches
‘selecting a subset of at least one general set in meeting an application desire
Appeal 2010-011109
Application 11/259,597

6
for the vocabulary of color names,’” and does not teach or suggest that the
“selecting” is performed according to a rate or standard that would be
necessary to constitute “scaling” as disclosed and claimed by Appellant. Id.
(citing FOA 3 and Mojsilovic ¶ [0082]).
In addition, Appellant argues that their Specification defines “scaling”
as being accomplished by scaling component 228 which specifies a subset of
the set of lexical classifiers from which lexical classifiers may be assigned
for a given application and that their color naming system may be scaled to
assign lexical classifiers from a large number of names or a small number of
names, depending on the intended application. Further, Appellant states that
scaling component 228 may operate by adding names in terms of relative
frequency of use, i.e., by a “rate or standard.” App. Br. 10-11 (citing Spec.
¶ [0020]).
In particular, Appellant contends that Beretta fails to teach or suggest
a “database comprising thousands of lexical classifiers given by human
subjects to different visual attribute values for image elements,” as recited in
claim 1, because Beretta, like Mojsilovic, teaches reliance upon government
standards and dictionaries. App. Br. 12-13.
Finally, Appellant contends that the differences between the cited
prior art and claim 1 are significant because neither Mojsilovic and Beretta
“provide[s] the ability to scale a large database to a smaller subset, which
may allow the searched set to be limited to a specific range for a certain
application, either through user specified directions or a scaling algorithm.”
App. Br. 13.
Appeal 2010-011109
Application 11/259,597

7
In response to Appellant’s contentions regarding Issue 1(a), the
Examiner finds that the process described in Appellant’s Specification and
relied upon by Appellant to distinguish the recited scaling operation over the
cited art allows computer implemented algorithms or user input to perform
scaling or selecting, which Appellant defines as selecting a subset of a large
database for further action in order to limit search times and target interest
areas. Ans. 14-15 (citing Spec. p. 6, ll. 11-27). The Examiner analogizes
this process as being similar to that which a person would undertake when
selecting a paint color at a paint store for a painting project, i.e., there is a
large array of color chips and segments in which one or more color ranges or
hues are selected to choose “the best color” for the project. Ans. 15.
We agree with the Examiner on this point and, in support of this
finding, we refer to the Examiner’s citation of Mojsilovic as teaching subset
scaling or selection, and Mojsilovic’s teaching of “color names” as
corresponding to the recited “lexical classifiers.” App. Br. 15 (citing
Mojsilovic ¶ [0082]; and claims 2, 10, and 30).
Further, with respect to the recitation of “said database comprising
thousands of lexical classifiers given by human subjects to different visual
attribute values for image elements,” we also agree with the Examiner’s
finding that the combination of Beretta and Mojsilovic teach this limitation
(Ans. 16-17). Beretta describes a large color database compiled at the
National Bureau of Standards (NBS) (see Beretta col. 24:59 through col.
25:15, and col. 53:3-6 (citing U.S. Dept. of Commerce, National Bureau of
Standards in Color: Universal Language and Dictionary Names, NBSS
Special Publication 440, Washington, D.C., 1976). In addition, we agree
Appeal 2010-011109
Application 11/259,597

8
with the Examiner (Ans. 17) that Mojsilovic teaches using a subset of a
number of most-used color names contained in a database. See, Mojsilovic
¶ [0007] (citing to a known color listing by Maerz and Paul, a known color
vocabulary by Munsell, as well as various versions of NBS color databases).
The disputed limitation of claim 1 recites, “scaling a database
including a set of lexical classifiers corresponding to a particular type of
visual attribute to a smaller subset of said database based on an intended
application, said database comprising thousands of lexical classifiers given
by human subjects to different visual attribute values for image elements.”
Claim 1. We are not persuaded by Appellant’s argument (App. Br. 11;
Reply Br. 4) that “[t]he prior art does not teach or suggest that this
‘selecting’ is performed according to a rate or standard as would be required
for the action to constitute ‘scaling’ as disclosed and claimed by the
Appellant.” App. Br. 11.
During examination, a claim must be given its broadest reasonable
interpretation consistent with the Specification, as one of ordinary skill in
the art would interpret it. Because the applicant has the opportunity to
amend claims during prosecution, giving a claim its broadest reasonable
interpretation will reduce the possibility that the claim, once issued, will be
interpreted more broadly than is justified. In re Yamamoto, 740 F.2d 1569,
1571 (Fed. Cir. 1984); In re Zletz, 893 F.2d 319, 321 (Fed. Cir. 1989)
(“During patent examination the pending claims must be interpreted as
broadly as their terms reasonably allow.”).
Appeal 2010-011109
Application 11/259,597

9
Despite Appellant’s disclosure of the optional use of a rate or
standard,3 Appellant’s argument is not commensurate with the scope of the
claim, since the claim does not require that the recited “scaling” be based
upon a rate or standard.4 We therefore find that the combination of art cited
by the Examiner teaches or suggests the limitation in dispute for the reasons
discussed above.
With respect to Appellant’s argument of improper motivation to
combine Mojsilovic with Beretta, i.e., Issue 1(b), Appellant contends claim 1
recites the scaled database being a “database comprising thousands of lexical
classifiers given by human subjects to different visual attribute values for
image elements,” but that Mojsilovic teaches away from the reference
combination because it refers instead to a plurality of different databases or
dictionaries of available colors, and not to lexical classifiers given by human
subjects. App. Br. 11-12.
“A reference may be said to teach away when a person of ordinary
skill, upon reading the reference, would be discouraged from following the
path set out in the reference, or would be led in a direction divergent from
the path that was taken by the applicant.” In re Gurley, 27 F.3d 551, 553

3 See Spec. p. 6, ll. 17-20 (“[t]he scaling component 228 may operate
algorithmically, that is, by adding the names in terms of relative frequency
of use or by using less commonly used names later.”) (emphasis added).
4 Appellant also relies upon a dictionary definition of “scaling” as being “to
pattern, make, regulate, set, or estimate according to some rate or standard.”
Reply Br. 4 (citing http://www.merriam-webster.com/dictionary/scaling.)
(emphasis in original). We are similarly unpersuaded by reliance upon this
definition in construing the claims under the broadest reasonable
interpretation standard, cited supra.
Appeal 2010-011109
Application 11/259,597

10
(Fed. Cir. 1994); Para-Ordnance Mfg., Inc. v. SGS Importers Int’l., Inc., 73
F.3d 1085, 1090 (Fed. Cir. 1995).
In this case, Mojsilovic does not lead to a divergent direction from the
claimed subject matter or teach away from the claimed invention but, in
direct contradiction to Appellant’s argument, the reference instead directly
teaches the limitation in dispute because the databases cited in the prior art
were compiled by humans, and the reference also teaches use of a subset
containing the most commonly used names.
Accordingly, Appellant has not provided sufficient evidence or
arguments to persuade us of any reversible error in the Examiner’s
characterization of the cited art and related claim construction. Therefore,
we sustain the Examiner’s unpatentability rejection of independent claim 1.
As Appellant has not provided separate arguments with respect to dependent
claims 2-5 and 19-24, we similarly sustain the Examiner’s rejection of these
claims under 35 U.S.C. § 103(a).
2. Rejection of Claims 7-11 and 13-17
Issue 2
Appellant argues (App. Br. 14-16; Reply Br. 8-11) that the
Examiner’s rejection of claim 7 under 35 U.S.C. § 103(a) as being
unpatentable over the combination of Mojsilovic and Prabhakar is in error.
These contentions present us with the following issue:
Did the Examiner err in finding that Appellant’s claimed
system is unpatentable because the combination of Mojsilovic
and Prabhakar teaches or suggests “a processor for . . .
Appeal 2010-011109
Application 11/259,597

11
generating a histogram for each lexical classifier in a specified
vocabulary based on usage of that lexical classifier in said
database; and generating a set of fuzzy membership functions
based on said histograms; wherein said designated lexical
classifier is produced by said processor in accordance with said
fuzzy membership functions,” as recited in claim 7?
Analysis
Appellant contends that Mojsilovic discloses processing and display
of a color name histogram, “but does not teach or suggest that the histogram
may be used in assigning a lexical classifier to an initial attribute value.”
App. Br. 15 (citing Mojsilovic ¶ [0078] and Fig. 15). Appellant also
contends that Prabhakar discloses using a “soft” fuzzy image classification
to determine whether an input image area is in a graphics or picture class,
“but does not teach or suggest generating fuzzy membership functions based
on histograms of lexical classifiers based on the usage of the lexical
classifiers in a database.” App. Br. 15. Appellant also contends that the
differences between the cited prior art and claimed subject matter are
significant because Appellant’s system allows a machine learning algorithm
to transform an initial attribute value from an input image to a designated
lexical classifier using fuzzy membership functions based on histograms of
lexical classifier usage, such that the rejection of independent claim 7 over
the combination of Mojsilovic and Prabhakar should be withdrawn.
In response, the Examiner finds that Mojsilovic teaches the first
limitation, i.e., “generating a histogram for each lexical classifier in a
specified vocabulary based on usage of that lexical classifier in said
database” (Ans. 18 (citing Mojsilovic ¶ [ 0078]), and Prabhakar teaches the
Appeal 2010-011109
Application 11/259,597

12
second group of limitations, i.e., “generating a set of fuzzy membership
functions based on said histograms; wherein said designated lexical
classifier is produced by said processor in accordance said fuzzy
membership functions,” as recited in claim 7. Ans. 18 (citing Prabhakar
¶¶ [0068]-[0071]).
We agree with the Examiner’s findings regarding the teachings of
Mojsilovic because the portion cited by the Examiner teaches that data
processor 101 computes a color name based on its color value for each
perceptually significant image pixel, and a set of rules that describe different
color naming patterns in humans is loaded. Data processor 101 then
computes the histogram of color names for each accuracy level. See
Mojsilovic ¶ [0078] and Fig. 14 (steps 1403-1405). We find that at least this
portion of Mojsilovic teaches the recited “generating a histogram for each
lexical classifier in a specified vocabulary based on usage of that lexical
classifier in said database.” Appellants repeated arguments concerning this
point in the Reply Brief are not persuasive. Reply Br. 8-10.
We also agree with the Examiner’s findings regarding the teachings of
Prabhakar because the cited portion teaches the concept of fuzzy logic
implemented in an image processing system that uses a “soft” image
classification process in which image areas are classified in one of a picture,
graphics, or fuzzy class based upon use of image histogram information.
Prabhakar ¶¶ [0068]-[0071]; see also ¶¶ [0050]-[0052], and Fig. 3.
In addition to the general concept of the use of fuzzy classifiers
determined by histograms as taught by Prabhakar, we note that Mojsilovic,
which is also directed to Appellant’s specific technical field of visual
Appeal 2010-011109
Application 11/259,597

13
attribute management (e.g., color management), also teaches the use of
similar fuzzy logic and set theory. Mojsilovic ¶ [0005], line 40 et seq. For
example, Mojsilovic teaches in its Background section that the concept of
graded or fuzzy membership in color categories is known in the art; that
prior art color models have considered at least four fuzzy sets, i.e., red,
green, yellow, and blue; and that supporting other color terms requires the
introduction of new and ad hoc fuzzy set operations. Id.
Accordingly, Appellant has not provided sufficient evidence or
arguments to persuade us of any reversible error in the Examiner’s
characterization of the cited art and related claim construction. Therefore,
we sustain the Examiner’s unpatentability rejection of independent claim 7.
As Appellant has not provided separate arguments with respect to
independent claim 13 or dependent claims 8-11 and 14-17, depending from
claims 7 and 13, respectively, we similarly sustain the Examiner’s rejection
of these claims under 35 U.S.C. § 103(a).
CONCLUSION
The Examiner did not err with respect to the various unpatentability
rejections of claims 1-5, 7-11, 13-17, and 19-24 over the prior art of record,
and the rejections are sustained.
DECISION
The decision of the Examiner to reject claims 1-5, 7-11, 13-17, and
19-24 is affirmed.
Appeal 2010-011109
Application 11/259,597

14
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(1)(iv) (2011).
AFFIRMED

ELD