Ex Parte Fradkin et al

Board of Patent Appeals and Interferences

Jul 5, 2012

10535466 (B.P.A.I. Jul. 5, 2012)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte MAXIM FRADKIN, JEAN-MICHEL ROUET,
and SHERIF MAKRAM-EBEID
____________

Appeal 2010-002792
Application 10/535,466
Technology Center 2400
____________

Before JOHN A. JEFFERY, THOMAS S. HAHN, and
BRUCE R. WINSOR, Administrative Patent Judges.

JEFFERY, Administrative Patent Judge.

DECISION ON APPEAL
Appeal 2010-002792
Application 10/535,466

2
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1-15. We have jurisdiction under 35 U.S.C. § 6(b). We
affirm.

STATEMENT OF THE CASE
Appellants’ invention automatically adapts a 3D mesh model onto an
object’s 3D surface by, among other things, dynamically adapting model
resolution to image features. See generally Abstract. Claim 1 is illustrative
with key disputed limitations emphasized:
1. An image processing system having image data processing means
of automatic adaptation of 3-D Mesh Model to image features, for Model-
based image segmentation, comprising means of dynamic adaptation of the
Model resolution to image features including means of locally setting higher
resolution when reliable image features are found and means of setting
lower resolution in the opposite case; and comprising viewing means for
visualizing the images.

THE REJECTIONS
1. The Examiner rejected claims 1, 2, and 15 under 35 U.S.C.
§ 103(a) as unpatentable over Hervé Delingette, General Object
Reconstruction Based on Simplex Meshes, 32 INT’L J. COMP. VISION 111-46
(1999). Ans. 3-5.1
2. The Examiner rejected claims 3-7 under 35 U.S.C. § 103(a) as
unpatentable over Delingette and Bernardini (US 6,968,299 B1; Nov. 22,
2005 (filed Apr. 14, 2000)). Ans. 6-8.

1 Throughout this opinion, we refer to (1) the Appeal Brief filed June 18,
2009; (2) the Examiner’s Answer mailed September 10, 2009; and (3) the
Reply Brief filed November 5, 2009.
Appeal 2010-002792
Application 10/535,466

3
3. The Examiner rejected claims 8-14 under 35 U.S.C. § 103(a) as
unpatentable over Delingette, Bernardini, and Vannah (US 6,201,889 B1;
Mar. 13, 2001). Ans. 8-12.

THE OBVIOUSNESS REJECTION OVER DELINGETTE
The Examiner finds that Delingette discloses every recited feature of
representative claim 1 except for expressly disclosing an image processing
system, but Delingette is said to nonetheless evidence using such a system in
concluding that the claim would have been obvious. Ans. 3-5.
Appellants argue that Delingette does not teach or suggest (1) means
of locally setting higher resolution when reliable image features are found,
and (2) means of setting lower resolution in the opposite case as claimed.
App. Br. 4-5; Reply Br. 2-4. Although Appellants acknowledge that
Delingette increases resolution in areas of high curvature, and that some
high curvature regions may include reliable image features, Appellants
nonetheless contend that Delingette does not determine whether to raise or
lower resolution based on reliable image features. Id.

ISSUE
Under § 103, has the Examiner erred in rejecting claim 1 by finding
that Delingette would have taught or suggested (1) means of locally setting
higher resolution when reliable image features are found, and (2) means of
setting lower resolution in the opposite case?

Appeal 2010-002792
Application 10/535,466

4
ANALYSIS
This dispute turns on one key question: What is a “reliable” image
feature? According to the Examiner, any image feature qualifies as
“reliable”—including high curvature values in a noisy image. Ans. 13.
Appellants, however, contend that a noisy image feature can be “unreliable”
despite having points of high curvature. App. Br. 5.
Appellants’ Specification does not define a “reliable” image feature,
but does note that model resolution is (1) raised when reliable image features
are found, and (2) lowered “in the opposite case, when image features are far
or are not reliable, for instance noisy.” Spec. 6:5-8 (emphasis added).
Although this example of an “unreliable” image feature is hardly limiting, it
nonetheless indicates that at least “noisy” image features are unreliable. The
Specification, however, does not specify what degree of noise—however
slight—is sufficient to render image features “noisy” and therefore
“unreliable” in Appellants’ parlance. Nor does Appellants’ Specification
specify what other attributes of image features render them “reliable” or
“unreliable,” let alone define these terms in the context of the claimed
invention.
But what we can infer from Appellants’ somewhat limited disclosure
in this regard is that a “noisy” image feature is “unreliable”—despite having
high curvature values. Accord App. Br. 5. Therefore, the Examiner’s
position to the contrary regarding the “extreme case”2 is somewhat
overstated, at least to the extent that “noisy” (and therefore “unreliable”)
image features are distinguishable from other image features that are not

2 See Ans. 13 (“In the extreme case, high curvature values in a noisy image
can be qualified as ‘reliable image features’.”).
Appeal 2010-002792
Application 10/535,466

5
“noisy.” In any event, when interpreted in light of the Specification, a
“reliable” image feature as claimed encompasses a wide variety of image
features—including high-curvature regions—so long as they are not “noisy.”
Turning to the prior art, Appellants admit that Delingette increases
resolution in high-curvature areas in Figure 9(b). App. Br. 4; Reply Br. 3.3
Accord Delingette, at 121 (captioning Figure 9(b) as “[d]eformable simplex
after adaptation of its reference metric parameter in order to increase the
resolution in places with high curvature” (emphasis added)). Appellants
also admit that some high curvature regions may include reliable image
features. Reply Br. 3. Based on these facts, skilled artisans would recognize
that, in a scenario involving noiseless images, Delingette increases
resolution when reliable image features are found, namely high curvature
regions. That Delingette’s high curvature regions may include unreliable
image features in other scenarios (e.g., involving noisy images) as
Appellants contend (App. Br. 4-5) does not preclude increasing resolution in
the scenario where reliable image features are found. Accord Ans. 13 (“The
fact that some high curvature regions are unreliable does not by itself
exclude high curvature regions from being reliable.”). Moreover, by
increasing resolution when reliable image features (i.e., areas of high
curvature) are found in Delingette, it follows that resolution is relatively
lower in the opposite case involving the non-high-curvature regions. See
Delingette, at 121; Fig. 9(b).

3 Although Appellants ambiguously indicate that “Delingette does or does
not increase resolution in areas of high curvature” in the fourth paragraph of
page 3 of the Reply Brief (emphasis added), Appellants nonetheless admit
that Delingette increases resolution in these areas in the first full paragraph
of page 3 of the Reply Brief (citing App. Br. 4).
Appeal 2010-002792
Application 10/535,466

6
That there are a finite number of identified, predictable solutions in
adapting resolution when finding reliable or unreliable image features as
summarized below only bolsters the Examiner’s obviousness position.

Resolution Image Feature
Increase Reliable
Decrease Unreliable
No change

Table 1: Summary of Identified Predictable Solutions for Setting
Resolution for Reliable or Unreliable Image Features

Where, as here, there is (1) a design need to solve a problem (i.e.,
changing resolution based on image features), and (2) a finite number of
identified, predictable solutions (i.e., the specific resolution and image
feature options noted above), ordinarily skilled artisans would have good
reason to pursue those options within their technical grasp. See KSR Int’l
Co. v. Teleflex Inc., 550 U.S. 398, 421 (2007). Accordingly, we find it
would have been obvious to try these resolution and image feature options, a
subset of which corresponds to the disputed limitations of the claimed
invention.
We are therefore not persuaded that the Examiner erred in rejecting
representative claim 1, and claims 2 and 15 not separately argued with
particularity.

Appeal 2010-002792
Application 10/535,466

7
THE OTHER OBVIOUSNESS REJECTIONS
We also sustain the Examiner’s obviousness rejections of claims 3-14.
Ans. 6-12. Despite nominally arguing these claims separately, Appellants
reiterate similar arguments made in connection with claim 1 and allege that
the additional cited references fail to cure those purported deficiencies. App.
Br. 5-9; Reply Br. 4-7. We are not persuaded by these arguments, however,
for the reasons previously discussed.

CONCLUSION
The Examiner did not err in rejecting claims 1-15 under § 103.

ORDER
The Examiner’s decision rejecting claims 1-15 is affirmed.
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).

AFFIRMED

babc