Ex Parte Herley

Board of Patent Appeals and Interferences

Aug 28, 2009

09970144 (B.P.A.I. Aug. 28, 2009)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte CORMAC E. HERLEY
_____________

Appeal 2006-003217
Application 09/970,144
Technology Center 2600
____________

Decided: August 28, 2009
____________

Before KENNETH W. HAIRSTON, ROBERT E. NAPPI,
and ELENI MANTIS MERCADER, Administrative Patent Judges.

MANTIS MERCADER, Administrative Patent Judge.

DECISION ON APPEAL

Appeal 2006-003217
Application 09/970,144

2
STATEMENT OF THE CASE
Appellant seeks our review under 35 U.S.C. § 134(a) of the Examiner’s final
rejection of claims 1-8, 11-20, 22-29, and 31-33. We have jurisdiction under 35
U.S.C. § 6(b).
We affirm.

INVENTION
Appellant’s claimed invention is directed to employing a digital imaging
device camera with VGA resolution employing a multi-frame reconstruction
algorithm which can achieve a 4x improvement in resolution in each dimension
(Spec. 2: ¶ [0005]). The multi-frame reconstruction entails mapping non-uniform
samples from multiple images to a uniform or regular grid (Spec. 9: ¶ [0038]).
Claim 1, reproduced below, is representative of the subject matter on appeal:
1. A text document capture method for digitizing a text document segment
in printed form, comprising:
imparting a continuous lateral jittering between a digital imaging device and
the text document;
obtaining multiple laterally-displaced digital images of all of the text
document segment during the continuous lateral jittering and determining
fractional pixel offset positions at which each image was obtained;
forming from the multiple laterally displaced images an enhanced resolution
representation of the text document as a function of the fractional pixel offset
positions;
and
de-blurring the enhanced resolution representation of the text document by
thresholding the enhanced resolution representation into either one of two pixel
luminance levels, representing foreground and background pixels, with the
foreground pixels corresponding to text in the text document.

Appeal 2006-003217
Application 09/970,144

3
THE REJECTION
The Examiner relies upon the following as evidence of unpatentability:
Crinon US 6,285,804 B1 Sep. 04, 2001
(filed Dec. 21, 1998)

Steinkirchner US 5,392,365 Feb. 21, 1995

The following rejection is before us for review:
The Examiner rejected claims 1-8, 11-20, 22-29, and 31-33 under 35 U.S.C.
§ 103(a) as being unpatentable over Crinon in view of Steinkirchner.
Appellant argues the art rejection of claims 1-8, 11-20, 22-29, and 31-33 as a
group with independent claims 1, 13, 25, and 33 as representative (App. Br. 9-15).1
Although Appellant nominally argues independent claims 13, 25, and 33 separately
(App. Br. 12-15), these arguments merely reiterate the same arguments presented
for claim 1. Accordingly, we group claims 1-8, 11-20, 22-29, and 31-33 together
and select claim 1 as representative. Thus, claims 2-8, 11-20, 22-29, and 31-33,
stand or fall with claim 1. See 37 C.F.R. § 41.37 (c)(1)(vii) (2004). We note that
the Examiner withdrew the rejections made under 35 U.S.C. § 112, second
paragraph (Ans. 12), and, thus, we do not address Appellant’s arguments with
respect to these rejections (App. Br. 6-8).

1 Only arguments made by Appellant have been considered in this decision.
Arguments which Appellant could have made but did not make in the Brief have
not been considered and are deemed waived. See 37 C.F.R. § 41.37(c)(1)(vii)
(2004).
Appeal 2006-003217
Application 09/970,144

4
OBVIOUSNESS
ISSUE
Appellant admits that Crinon, much like Appellant’s disclosed method,
teaches a method that makes use of a plurality of images having potential offsets
measured in fractional pixel instances (App. Br. 10). Appellant however asserts
that in stark contrast to the Crinon reference, rather than computing global motion
models, deriving motion vectors from the motion model for relating a reference
image to every other image, mapping image pixels to “‘inter-pixel positions’”
which are in turn used to map pixel intensities to “‘high-resolution grid points,’”
the Appellant simply computes enhanced resolution images as “a direct function
of the fractional pixel offset positions,” as described and claimed (App. Br. 10).
Appellant asserts that the construction of high resolution images is based on
fractional pixel offsets “without the need to compute the motion models” as
disclosed in Crinon (Reply Br. 5). Appellant explains that the function includes a
weighted combination of multiple image pixel samples that are nearest the
enhanced resolution sample (App. Br. 10).
The Examiner responds that based on Appellant’s arguments stating that
Crinon teaches mapping image pixels to “‘inter pixel positions’” which are in turn
used to map pixel intensities to “‘high resolution grid points,’” Appellant does not
dispute that Crinon creates an enhanced resolution image from low resolution
images “‘based on’” fractional pixel offset positions (Ans. 12-13). The Examiner
interpreted the phrase “as a function” to be “‘based on’” and not to mean an
equation (Ans. 12-13), wherein the equation is the weighted sum of the image
values of the nearest image pixel values as argued by Appellant (App. Br. 10). The
Appeal 2006-003217
Application 09/970,144

5
Examiner nonetheless states that Crinon teaches that four nearest pixel values are
taken and the fractional pixel offset of each is used to determine the weighting
factor (Ans. 13-14).
The issue, then, is whether Appellant has shown that the Examiner erred by
finding that Crinon teaches “forming . . . an enhanced resolution representation
. . . as a function of the fractional pixel offset positions” as recited in claim 1.

FINDINGS OF FACT
The following relevant findings of fact (FF) are supported by a
preponderance of the evidence:
1. Independent claims 1, 13, 25, and 33 only require “a function of the
fractional pixel offset positions” and the claim language is silent as to
whether this is a direct or indirect function (claims 1, 13, 25, and 33). The
claims use the transitional term “comprising” (claims 1, 13, 25, and 33).
2. Crinon teaches deriving motion vectors 22 from a parametric global motion
model (col. 3, ll. 53-58)
3. Appellant admits that Crinon teaches mapping image pixels to “‘inter-pixel
positions’” which are in turn used to map pixel intensities to “‘high-
resolution grid points’” (App. Br. 10), and that Crinon’s method includes a
plurality of images having potential offsets measured in fractional pixel
distances (App. Br. 10).
4. Crinon teaches that the interpolation process (i.e., function) makes use of the
associated distance values of the closest pixels (i.e., four pixels 24 shown in

Appeal 2006-003217
Application 09/970,144

6
the 4 distinct squares at the right section of Fig. 5) to calculate the sample
value at the high resolution grid point 20 (col. 5, ll.45-49).
5. Crinon teaches that the interpolation process includes a weighing factor
specifying the contribution of the associated sample value in the final image
value at grid point 20 (col. 5, ll. 49-53).
6. Crinon further teaches that the motion vector 22(2) mapped to the image
16B at time t=to+2 has an inter-pixel position 24B closest to the existing
pixel position 26B (col. 5, ll. 53-55).
7. Crinon teaches that the distance di between position 24B and 26B is the
smallest distance for any one of the motion vectors 22 (i.e., 4 motion vectors
shown in Fig. 5) at high resolution point 20A (col. 5, ll. 55-58).
8. Crinon teaches that the pixel intensity value for grid location 20A is derived
using spatial interpolation (col. 5, ll. 58-59).

PRINCIPLES OF LAW
The Examiner bears the initial burden of presenting a prima facie case of
obviousness, and Appellant has the burden of presenting a rebuttal to the prima
facie case. In re Oetiker, 977 F.2d 1443, 1445 (Fed. Cir. 1992). Appellant has the
burden on appeal to the Board to demonstrate error in the Examiner’s position. See
Kahn, 441 F.3d 977, 985-86 (Fed. Cir. 2006).
The transitional term “comprising” is inclusive or open-ended and does not
exclude additional, unrecited elements. Genentech, Inc. v. Chiron Corp., 112 F.3d
495, 501 (Fed. Cir. 1997).

Appeal 2006-003217
Application 09/970,144

7
Although claims are interpreted in light of the specification, limitations from
the specification are not read into the claims. In re Van Geuns, 988 F.2d 1181,
1184 (Fed. Cir. 1993).

ANALYSIS
At the outset, we note that Appellant’s assertion (App. Br. 10) that the
claims require “a direct function of the fractional pixel offset positions,” is
misplaced as the claim language only requires “a function of the fractional pixel
offset positions” and the claim language is silent as to whether this is a direct or
indirect function (FF 1). Thus, the claim language does not preclude an indirect
function for determining the factional pixel positions. Furthermore, the additional
features of computing global motion models and deriving motion vectors from the
motion model for relating a reference image to every other image as taught by
Crinon (FF 2) are not precluded because the transitional term “comprising” is
inclusive or open-ended and does not exclude additional, unrecited elements (FF
1). See Genentech, 112 F.3d at 501.
Appellant admits that Crinon teaches mapping image pixels to “‘inter-pixel
positions’” which are in turn used to map pixel intensities to “‘high-resolution grid
points’” and that Crinon’s method includes a plurality of images having potential
offsets measured in fractional pixel distances (FF 3). Thus, Appellant admits that
Crinon teaches “forming . . . an enhanced resolution representation” (i.e., mapping
pixels to the high-resolution grid) “as a function of the fractional pixel offset
positions” (i.e., by using the potential offsets measured in fractional distances).

Appeal 2006-003217
Application 09/970,144

8
Appellant explains that the function includes a weighted combination of
multiple image pixel samples that are nearest the enhanced resolution pixel (App.
Br. 10). We note that the claim does not recite “a weighted combination of
multiple (e.g., 3) image pixel samples that are nearest the enhanced resolution
pixel.” So even if “function” meant “equation” the scope of the claim is not
limited to the specifics recited in the Specification as Appellant chose not to claim
the specifics. In other words, although claims are interpreted in light of the
specification, limitations from the specification are not read into the claims. See
Van Geuns, 988 F.2d at 1184 (Fed. Cir. 1993).
Furthermore, while we agree with the Examiner that under the broadest
reasonable interpretation “as a function” means “‘based on’” (Ans. 12-13) and not
“an equation,” nonetheless Crinon teaches what Appellant describes as a function
as will be illustrated infra.2
Crinon teaches that the interpolation process (i.e., function) makes use of the
associated distance values of the closest pixels (i.e., four pixels 24 shown in the 4
distinct squares at the right section of Fig. 5) to calculate the sample value at the
high resolution grid point 20 (FF 4). Crinon teaches that the interpolation function
includes a weighing factor specifying the contribution of the associated sample
value in the final image value at grid point 20 (FF 5). Crinon further teaches that
the motion vector 22(2) mapped to the image 16B at time t=to+2 has an inter-pixel
position 24B closest to the existing pixel position 26B (FF 6). Crinon teaches that

2 We note that while Appellant states that “as a function” means “based on” (Reply
Br. 6) and not an equation, Appellant nonetheless focuses on the calculated
weighted sum (i.e., an equation) (Reply Br. 5-6).
Appeal 2006-003217
Application 09/970,144

9
the distance di between position 24B and 26B is the smallest distance for any one
of the motion vectors 22 (i.e., 4 motion vectors shown in Fig. 5) at high resolution
point 20A (FF 7). Crinon teaches that the pixel intensity value for grid location
20A is derived using spatial interpolation (FF 7). Accordingly, Crinon teaches an
interpolation function that includes a weighted combination (i.e., weighing factor
specifying the contribution of the associated sample value in the final image value
at grid point 20) of multiple pixels (i.e., four pixels 24 shown in Fig. 5) that are
nearest to the enhanced resolution sample 20A.
For the foregoing reasons, Appellant has not persuaded us that the Examiner
erred in rejecting claim 1 and claims 2-8, 11-20, 22-29, and 31-33 which fall with
claim 1.

CONCLUSION
Under 35 U.S.C. § 103, Appellant has not shown that the Examiner erred by
finding that Crinon teaches “forming . . . an enhanced resolution representation
. . . as a function of the fractional pixel offset positions” as recited in claim 1, and
similarly recited in claims 13, 25, and 33. Similarly, Appellant has not persuaded
us that the Examiner erred in rejecting claims 2-8, 11-12, 14-20, 22-24, 26-29, and
31-32 which fall with claim 1.

ORDER
The decision of the Examiner to reject claims 1-8, 11-20, 22-29, and
31-33 under 35 U.S.C. § 103, is affirmed.
Appeal 2006-003217
Application 09/970,144

10
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

KIS

MICROSOFT CORPORATION
C/O LYON & HARR, LLP
300 ESPLANADE DRIVE
SUITE 800
OXNARD, CA 93036