Ex Parte WU et al

Patent Trial and Appeal Board

Aug 30, 2016

13166356 (P.T.A.B. Aug. 30, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
13/166,356 06/22/2011
12716 7590 09/01/2016
Marshall, Gerstein & Bornn LLP (Google)
233 South Wacker Drive
6300 Willis Tower
Chicago, IL 60606-6357
FIRST NAMED INVENTOR
XiaqingWU
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www .uspto.gov
ATTORNEY DOCKET NO. CONFIRMATION NO.
31730/3984-00 5718
EXAMINER
GILLIARD, DELOMIA L
ART UNIT PAPER NUMBER
2665
NOTIFICATION DATE DELIVERY MODE
09/01/2016 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address( es):
mgbdocket@marshallip.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte XIAQING WU and LIN YANG
Appeal2015-003299
Application 13/166,356
Technology Center 2600
Before CARL W. WHITEHEAD JR., HUNG H. BUI, and
JON M. JURGOV AN, Administrative Patent Judges.
JURGOV AN, Administrative Patent Judge.
DECISION ON APPEAL
Appellants 1 seek review under 35 U.S.C. § 134(a) from a Final
Rejection of claims 1-15. We have jurisdiction under 35 U.S.C. § 6(b).
We affirm-in-part. 2
1 Appellants identify Google, Inc. as the real party in interest. (App. Br. 3.)
2 Our Decision refers to the Specification filed June 22, 2011 ("Spec."), the
Final Office Action mailed June 18, 2013 ("Final Act."), the Appeal Brief
filed July 1, 2014 ("App. Br."), the Examiner's Answer mailed Dec. 26,
2014 ("Ans."), and the Reply Brief filed Feb. 26, 2015 ("Reply Br.").
Appeal2015-003299
Application 13/166,356
CLAIMED SUBJECT MATTER
The claims are directed to assigning tree species to clusters of tree
crowns in aerial imagery. (Spec. Abstract.) Claim 1, reproduced below, is
illustrative of the claimed subject matter:
1. A computer-implemented method, comprising:
(a) receiving an image of a portion of the Earth;
(b) detecting one or more tree crowns in the image;
( c) for each of the one or more tree crowns:
(i) representing the respective tree crown as a color,
texture and entropy histogram;
(ii) reducing the dimensionality of the color, texture
and entropy histogram to produce a reduced
dimensionality color, texture, and entropy histogram;
( d) clustering, into one or more clusters, similar reduced
dimensionality color, texture and entropy histograms for the one
or more tree crowns; and
( e) assigning each cluster to a tree species to match each
detected tree crown to a corresponding tree species.
(App. Br. 19- Claims App'x.)
REJECTIONS
Claims 1-6 and 8-14 stand rejected under 35 U.S.C. § 103(a) based
on Yang (Lin Yang et al., Tree Detection from Aerial Imagery, ACM GIS
'09, November 4--6, 2009), Tan (US 2009/0154795 Al, Jun. 18, 2009), and
Rousselle (US 7,212,670 Bl, May 1, 2007). (Final Act. 2-10.)
Claims 7 and 15 stand rejected under 35 U.S.C. § 103(a) based on
Yang, Tan, Rousselle, and Korber (Christian Korber et al., Multi-Channel
Texture Classification Applied to Feature Extraction in Forestry, ACM GIS
'09, November 4--6, 2009). (Final Act. 10-11.)
2
Appeal2015-003299
Application 13/166,356
ANALYSIS
Claims 1, 8, 10
A. Arguments Concerning Reducing Dimensionality and Clustering
Appellants argue the combined references fail to teach or suggest the
claim limitation "reducing the dimensionality of the color, texture and
entropy histogram [for each tree crown] to produce a reduced dimensionality
color, texture, and entropy histogram" as recited in claim 1, and
commensurately recited in claims 8 and 10. (App. Br. 13-15, Reply Br. 1-
3.) We do not find Appellants' argument persuasive.
As the Examiner finds (Final Act. 4--6, Ans. 13), Yang teaches use of
color, texture, and entropy with regard to tree detection in images. (Yang p.
132, Sect. 2.1 Feature selection.) Yang also teaches locating individual trees
from aerial images, and providing estimates of their crown size. (Ans. 13
citing Yang Abstract, Introduction.) Moreover, Yang states "[a]fter the
descriptors are computed for all aerial images, Principal Component
Analysis (PCA) is applied to reduce the dimensionality of descriptors to 12
while preserving about 95% of the total variance of the original descriptors."
(Yang p. 135, Sect. 5 Tree Detection for the Globe.) The Examiner finds
that substituting Tan's color, texture, and entropy histograms for Yang's
descriptors and applying Yang's dimension reduction yields the claimed
invention. (Final Act. 5 citing Tan i-fi-141--45.) We agree with the Examiner's
findings. See KSR Int'! Co. v. Teleflex Inc., 550 U.S. 398, 417 (2007)
3
Appeal2015-003299
Application 13/166,356
("KSR") (the simple substitution of one known element for another is
obvious.)
Appellants argue Yang reduces dimensionality for the entire aerial
image, and not just for each tree crown in the image, which is only part of an
image. (App. Br. 14.) There are several deficiencies in this argument. The
claims do not recite that each tree crown is only part of an image, so
Appellants' argument in this respect is not commensurate in scope with the
claim. See In re Self, 671F.2d1344, 1348 (CCPA 1982) (stating that
limitations not appearing in the claims cannot be relied upon for
patentability). Furthermore, the claims do not exclude the possibility that an
aerial image at a particular zoom level may be of only one tree crown. In
this case, reducing dimensionality of the entire image would be the same as
reducing dimensionality of the single tree crown. Furthermore, applying
Yang's PCA to the entire image reduces dimensionality of every object in
the image, including each tree crown. For all of these reasons, the claim
language fails to distinguish over the combined references.
In addition, Appellants argue the combined references fail to teach or
suggest the claim limitation "clustering, into one or more clusters, similar
reduced dimensionality color, texture, and entropy histograms for the one or
more tree crowns." (App. Br. 13-15.) The Examiner finds this limitation is
taught by Yang's description of clustering large-scale aerial imagery with
similar geographic features. (Final Act. 4 citing Yang p. 135, Sect. 5.)
Appellants' argument with respect to Yang is similar to that
previously discussed regarding reducing dimensionality, and asserts that
Yang's clustering is not for each tree crown, but for the aerial image as a
whole. As noted, Yang teaches locating individual trees in an aerial image,
4
Appeal2015-003299
Application 13/166,356
suggesting reduction in dimensionality may be performed for individual
trees. (Final Act. 4, Ans. 13-15 citing Yang, p. 131, Introduction.) The
Examiner thus has provided "some articulated reasoning with some rational
underpinning to support the legal conclusion of obviousness." KSR at 418
(2007) (quoting In re Kahn, 441 F.3d 977, 988 (Fed. Cir. 2006).
Furthermore, Appellants fail to consider the possibility that an image at a
particular zoom level may contain only one or no tree crowns, in which case
the claimed clustering is indistinguishable from Yang's teachings. In
addition, because performing clustering on the aerial image as a whole
includes each tree crown within that image, we are not persuaded by
Appellants' argument that the claim language distinguishes over the
combined references.
B. Argument that Combination of References Renders Yang
Unsatisfactory for Its Intended Purpose
Appellants further argue reducing dimensionality from an aerial image
to a tree crown would render Yang unsatisfactory for its intended purpose.
(App. Br. 15-16.) Specifically, Appellants argue applying Yang's
dimensionality reduction to tree crowns, not to the image as a whole, would
decrease efficiency of image processing, when Yang purports to achieve
improved efficiency by processing large portions of aerial imagery. (Id.
citing Yang p. 135, Sect. 5.)
If a proposed combination would render the prior art invention being
modified unsatisfactory for its intended purpose, then there is no suggestion
or motivation to make the proposed modification. In re Gordon, 733 F.2d
900 (Fed. Cir. 1984). However, Appellants provide no evidence to support
their assertion that applying Yang's dimensionality reduction to tree crowns,
5
Appeal2015-003299
Application 13/166,356
as opposed to the image as a whole, would decrease the efficiency of image
processing. Arguments of counsel cannot take the place of factually
supported objective evidence. In re Huang, 100 F.3d 135, 139--140 (Fed.
Cir. 1996).
Furthermore, as noted, Yang teaches locating individual trees from
aerial images, and providing estimates of their crown size. (Yang Abstract,
Introduction.) Accordingly, Yang does process imagery at the granularity of
tree crowns, so we do not agree Yang is limited to processing only entire
images, and not tree crowns therein.
Also, Appellants' argument again assumes an image in which there
are multiple tree crowns occupying respective parts of the image, whereas
the claim recites no such limitation. See In re Self, supra. For example, the
claims do not preclude there being a single tree crown occupying the entire
image at a particular zoom level, in which case processing the tree crown
may yield the same or greater efficiency as processing the entire image.
Furthermore, this argument fails to consider applying dimensionality
reduction to the entire image, in accordance with Yang's teachings, also
includes processing the tree crowns within the image.
For all of the foregoing reasons, we do not find this argument
persuasive to show Examiner error.
Claims 7 and 15
C. Rendering a Three-Dimensional Image
Appellants argue Korber fails to teach or suggest the claim limitation
"(f) receiving a desired view frustum" and "(g) rendering a three-
dimensional image according to the desired view frustum, wherein the
rendering includes one or more trees rendered based on an assigned tree
6
Appeal2015-003299
Application 13/166,356
species of a cluster" as recited in claim 7, and in commensurate language of
claim 15. (App. Br. 16-17 citing Korber Fig. 3, Sect. 2.2.) Appellants argue
Korber utilizes 3D graphics to explain a smoothing technique used to avoid
undesired edge effects, and that Korber' s Figure 3 is not a rendering based
on an assigned tree species cluster.
We agree with Appellants' argument. The cited portions of Korber,
relied upon in the rejection to teach this feature, do not mention receiving a
frustum or rendering a three-dimensional image. (See Final Act. 10 citing
Korber, Sects. 2.1, 2.2.) Accordingly, we do not sustain the rejection of
claims 7 and 15.
Claim 9
D. Tree Species Dictionary Storing Training Data
Claim 9 recites "(a) a receiver that receives one or more aerial images;
and (b) a tree species dictionary that stores training data." (App. Br. 20-
Claims App'x.) Appellants argue that Rousselle teaches a searchable
database, not a tree species dictionary storing training data, as claimed.
(App. Br. 17.)
The Examiner's rejection relies on Rousselle's teaching that "tree
crowns are analyzed by applying models which relate tree species data to
tree crown polygon area, and the data is stored in a searchable database, e.g.
as a vector file attribute." (Final Act. 8, 9, Ans. 18 citing Rousselle 4:18-
2[1], 6:3---6.)
We agree with the Examiner that Rousselle' s relation of tree species
data to tree crown area may be regarded as a tree species dictionary by a
7
Appeal2015-003299
Application 13/166,356
person of ordinary skill in the art, particularly in view of the lack of any
restrictive definition in the Specification or claims. 3
Furthermore, as the Examiner finds, Rousselle' s applied models may
be regarded as training data, particularly considering the lack of any
restrictive definition in the Specification or claims setting forth what the
training data is, or how it is used by any other element of the claim. 4
Applying a broad but reasonable interpretation to the claim terms consistent
with the Specification, the Examiner correctly finds Rousselle discloses the
claimed features. In re Am. Acad. of Sci. Tech. Ctr., 367 F.3d 1359, 1369
(Fed. Cir. 2004) (claims are given their broadest reasonable interpretation
consistent with the specification).
DECISION
We affirm the Examiner's rejection of claims 1-6 and 8-14 stand
rejected under 35 U.S.C. § 103(a).
We reverse the Examiner's rejection of claims 7 and 15 stand rejected
under 35 U.S.C. § 103(a).
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(l )(iv).
3 The Specification states various characteristics the tree species dictionary
may possess and thus does not define a tree species dictionary in a way that
distinguishes over Rouselles' teachings.
4 If there is further prosecution in this case, we advise the Examiner and
Appellants to consider whether claim 9 is sufficiently definite under 35
U.S.C. § 112, second paragraph, in view of the lack of any relationship of
the tree species dictionary or training data to any other element of the claim.
8
Appeal2015-003299
Application 13/166,356
AFFIRMED-IN-PART
9