Ex parte Keesen

Board of Patent Appeals and Interferences

Jan 31, 2000

08354929 (B.P.A.I. Jan. 31, 2000)

Application for patent filed December 12, 1994. 1
THIS OPINION WAS NOT WRITTEN FOR PUBLICATION
The opinion in support of the decision being entered today (1) was not written
for publication in a law journal and (2) is not binding precedent of the Board.
Paper No. 18
UNITED STATES PATENT AND TRADEMARK OFFICE
____________
BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________
Ex parte HEINZ-WERNER KEESEN
____________
Appeal No. 1997-2398
Application No. 08/354,9291
____________
ON BRIEF
____________
Before HAIRSTON, FLEMING, and BARRETT, Administrative Patent
Judges.
HAIRSTON, Administrative Patent Judge.
DECISION ON APPEAL
This is a decision on appeal under 35 U.S.C. § 134 from
the examiner’s final rejection of claims 1, 2, 8-12 and 22.
Claims 3-7, 13-21, 23 and 24 contain allowable subject matter
(Answer, page 1).
Appeal No. 1997-2398
Application No. 08/354,929
2
Briefly, and in general terms, the invention pertains to
a method and apparatus for estimating movement in an image
produced by a video signal. The inventors of the instant
application have recognized that movement estimation can be
performed using only specific bits, for example, one or more
most significant bits (MSB) of the picture element values of a
video image. Picture element values (pixels) in a reference
block of a first video image stored in a first storage unit
are compared with corresponding pixels in a search zone block
of a second video image stored in a second storage unit. The
comparison of corresponding pixels is performed by gates G1-
G6, counters C1-C6 and a comparator. In accordance with the
principles of the invention, the pixel comparison is performed
using fewer bits than the number nominally used to represent
each pixel. For example, pixels nominally represented by 8
bits are compared using only the one or two MSBs. Based on
this truncated bit width pixel comparison, the comparator
determines the search zone block and reference block that
gives the highest correlation (i.e., the blocks with the
greatest similarity). Upon identifying the search zone block
and reference block providing the highest correlation, the
Appeal No. 1997-2398
Application No. 08/354,929
3
comparator computes a movement vector representing the
relative position of the identified blocks.
Claim 1 is the only independent claim on appeal, and it
reads as follow:
1. A method of estimating movement in an image produced by a
video signal, comprising the steps of:
a) comparing picture element values in a reference block
of a first video image with corresponding picture element
values in a search zone block of a second video image, wherein
said comparison is performed using a number of bits N, less
than all of the bits used to represent a picture element
value; and
b) computing a movement vector using a relative position
of a matching search zone block to said reference block,
wherein said matching search zone block gives a highest
correlation of picture element values with the picture element
values of said reference block.
The prior art relied upon by the examiner as evidence of
obviousness are:
Music et al. (Music) 4,914,508 Apr. 3, 1990
Gobert et al. (Gobert) 5,247,586 Sep. 21, 1993
Claims 1, 2, 8-12 and 22 stand rejected under 35 U.S.C.
§ 103(a) as being unpatentable over Gobert in view of Music.
Rather than reiterate the examiner’s full statement of
the above-noted rejection and the conflicting viewpoints
advanced by the examiner and appellant regarding those
Appeal No. 1997-2398
Application No. 08/354,929
In our evaluation of the applied teachings, we have2
considered all of the disclosure of each teaching for what it
would have fairly taught one of ordinary skill in the art.
See In re Boe, 355 F.2d 961, 965, 148 USPQ 507, 510 (CCPA
1966). Additionally, this panel of the board has taken into
account not only the specific teachings, but also the
inferences which one skilled in the art would reasonably have
been expected to draw from the disclosure. See In re Preda,
401 F.2d 825, 826, 159 USPQ 342, 344 (CCPA 1968).
4
rejections, we make reference to the examiner’s answer (Paper
No. 15) for the examiner’s reasoning in support of the
rejection, and appellant’s main and reply briefs (Paper No. 14
and Paper No. 16, respectively) for appellant’s arguments
thereagainst.
Appellant has indicated that claims 1, 2, 8-12 and 22 all
stand or fall together (Brief, page 3). Accordingly, we
select claim 1 for review and shall decide the appeal on the
basis of this claim alone in keeping with 37 CFR §
1.192(c)(7).
OPINION
In reaching our decision in this appeal we have given
careful consideration to appellant’s specification and claims,
to the applied prior art references , and to the respective2
positions articulated by appellant and the examiner. As a
consequence of our review, we will not sustain the examiner’s
Appeal No. 1997-2398
Application No. 08/354,929
5
rejection of claims 1, 2, 8-12 and 22 under 35 U.S.C. §
103(a).
As a general proposition in an appeal involving a
rejection under 35 U.S.C. § 103, an examiner is under a burden
to make out a prima facie case of obviousness. If that burden
is met, the burden of going forward then shifts to the
applicant to overcome the prima facie case with argument
and/or evidence. Obviousness is then determined on the basis
of the evidence as a whole and the relative persuasiveness of
the arguments. See In re Oetiker, 977 F.2d 1443, 1445, 24
USPQ2d 1443, 1444 (Fed. Cir. 1992); In re Hedges, 783 F.2d
1038, 1039, 228 USPQ 685, 686 (Fed. Cir. 1986); In re
Piasecki, 745 F.2d 1468, 1472, 223 USPQ 785, 788 (Fed. Cir
1984); and In re Rinehart, 531 F.2d 1048, 1052, 189 USPQ 1147
(CCPA 1976).
We, of course, fully appreciate the examiner’s assessment
of the applied prior art, as well as the manner in which the
examiner proposes that the references be applied. However,
the difficulty that we have with the rejection advanced by the
examiner is that when we set aside what appellant has
disclosed to us in the present application, it is apparent to
Appeal No. 1997-2398
Application No. 08/354,929
6
us that the applied patents themselves would not have been
suggestive of the invention now claimed.
The examiner (Final Rejection, page 3) basically relies
on Gobert for teaching all the features required by
independent claim 1 except for the limitation of comparing
“... a number of bits N, less than all of the bits used to
represent a picture element value.” The examiner cites Music
as teaching the feature of “... a number of bits N, less than
all of the bits used to represent a picture element value,”
and contends that it would have been obvious to modify Gobert
by incorporating the bit-width truncation feature of Music
because such a modification would further reduce the
computational complexity goal expressed by Gobert (Final
Rejection, page 4). Further, the examiner states (Final
Rejection, pages 3 and 4) that Music suggests the
applicability of bit-width truncation in the area of frame-
differencing or motion compensation.
In rebuttal, the appellant requests reversal of the
examiner’s rejection asserting that the examiner’s rejection
misinterprets the Music and Gobert references and
misunderstands the effect of combining the features that these
Appeal No. 1997-2398
Application No. 08/354,929
7
two references disclose. In support thereof, the appellant
argues (Brief, page 4) that the use of truncated bit width
pixel comparison in motion estimation is not suggested by
Music or Gobert. Appellant contends (Brief, page 5) that,
unlike the instant invention defined by the claims, the bit-
width truncation taught by Music relates to RGB color
components for the expressed purpose of reducing the quantity
of data transmitted and improving data compression efficiency.
Appellant further asserts that neither Gobert nor Music
recognizes the advantage or the feasibility of using bit-width
truncation in motion estimation and neither reference supplies
any specific motivation for its use.
According to the appellant (Brief, page 7), the combination of
Gobert and Music would result in a system that produces an
additional computational error that deters the combination of
the Gobert and Music teachings.
In response to appellant’s first argument, the examiner
points to column 4, lines 10-35 of the disclosure of Music for
motivation for the combined teachings of Music with Gobert
(Answer, page 6). Further, the examiner asserts (Answer, page
7) that based on the teachings in column 14, lines 2-17 of
Appeal No. 1997-2398
Application No. 08/354,929
8
Music, "one of ordinary skill in the art would equate Music’s
'frame differencing advanced processing/compression scheme' as
executing 'motion estimation' as defined by Gobert."
We have reviewed the disclosure of Music, paying
particular attention to the passages relied upon by the
examiner, and we find that the teachings of Music would not
have been suggestive of using the disclosed bit-width
truncation feature in the field of motion estimation of video
signals. We are in general agreement with the appellant
(Brief, page 5) that the disclosure of Music is directed
toward digital color components and the compression of digital
color video data (Music, col. 2, line 34 to col. 3, line 45).
Music uses bit-width truncation of RGB color components (col.
4, lines 30-35) for compressing digital color data as opposed
to bit-width truncation for motion estimation of a video
image. We do not find any evidence here that would have been
suggestive of using bit-width truncation as it relates to
Gobert’s motion estimation. We are appreciative of the
examiner’s reliance on column 4, lines 10-35 of Music for its
teaching of reducing the amount of data required to represent
pixels of a picture. However, we find that this teaching is
Appeal No. 1997-2398
Application No. 08/354,929
9
the result of using run lengths in the video signal. Thus,
nothing in Music indicates that a reduction in hardware
complexity can be achieved as a result of using bit-width
truncation as the examiner has suggested (Answer, page 6).
The examiner points to the fact that both Gobert and
Music state that their inventions are relevant to the area of
video-conferencing (Answer, page 6). We find that this
acknowledgment would not have suggested the combination of the
teachings of Gobert and Music.
The examiner asserts (Final Rejection, page 4) that Music
suggests the applicability of bit-width truncation to motion
compensation, and, in support thereof, the examiner draws our
attention to column 14, line 60 to column 15, line 15 of
Music. We do not agree. To us, this portion of the Music
disclosure teaches the artisan that further compression of
color video data may be accomplished by the implementation of
these additional methods and in no way reveals to us that it
would have been obvious to one skilled in the art to use
Music’s bit-width truncation feature in the pixel comparison
method of Gobert for the purpose of estimating movement in an
image produced by a video signal.
Appeal No. 1997-2398
Application No. 08/354,929
10
Appellant’s argument (Brief, pages 4 and 8) emphasizes
that Gobert’s motion estimation involves significant error
since Gobert only selects certain pixels in correlating pixel
blocks instead of all of the pixels. Appellant points out
that Gobert’s method undermines the examiner's motivation for
additionally incorporating truncated bit-width pixel
comparison as taught by Music into the Gobert system because
this would compound the existing errors produced in the Gobert
method.
In rebuttal, the examiner disagrees with the appellant
noting that one possessing an ordinary level of skill in the
art would have been driven by reduction constraints to make
the combination of Gobert and Music. In fact, in the
examiner’s view (Answer, pages 12 and 13), the Gobert-Music
combination would have suggested several ways of reducing the
error in motion estimation and improving the accuracy of the
motion estimation. Firstly, as viewed by the examiner,
Gobert’s method includes the addition of more reference pixels
in the actual correlation process for improved accuracy.
Secondly, with the Gobert-Music combination, the examiner
Appeal No. 1997-2398
Application No. 08/354,929
11
reasons that the truncation amount could be varied to more
accurately reflect the pixel’s original component values.
We find ourselves in general agreement with the appellant
that the skilled artisan, equipped with Gobert's system and
objectives of designing a system including simplified hardware
implementation and reduced prediction error rate (col. 2,
lines 63-68), would not have been motivated to combine the
teachings of Music with the teachings of Gobert. More
specifically, we fail to see why the skilled artisan would
select certain pixel data as taught by Gobert (col. 1, lines
13-15), thereby incorporating a certain error rate into the
motion estimation of a video signal, and, thereafter, further
compound the error rate by truncating the bit data of the
pixel when it is generally understood that such truncation
would further increase the degree of error in the motion
estimation of a video signal. Accordingly, we will not
sustain the obviousness rejection of independent claim 1 and
dependent claims 2, 8-12 and 22. Therefore, the decision of
the examiner rejecting claims 1, 2, 8-12 and 22 is reversed.
DECISION
Appeal No. 1997-2398
Application No. 08/354,929
12
The decision of the examiner rejecting claims 1, 2, 8-12
and 22 under 35 U.S.C. § 103(a) is reversed.
REVERSED
KENNETH W. HAIRSTON )
Administrative Patent Judge )
)
)
)
) BOARD OF PATENT
LEE E. BARRETT ) APPEALS
Administrative Patent Judge ) AND
) INTERFERENCES
)
)
)
MICHAEL R. FLEMING )
Administrative Patent Judge )
lp
Appeal No. 1997-2398
Application No. 08/354,929
13
JOSEPH S TRIPOLI
PATENT OPERATION
GE AND RCA LICENSING MANAGEMENT
OPERATION INC CN 5312
PRINCETON NJ 08543-0028
Leticia
Appeal No. 1997-2398
Application No. 08/354,929
APJ HAIRSTON
APJ FLEMING
APJ BARRETT
DECISION: REVERSED
Send Reference(s): Yes No
or Translation (s)
Panel Change: Yes No
Index Sheet-2901 Rejection(s):
Prepared: October 13, 2000
Draft Final
3 MEM. CONF. Y N
OB/HD GAU
PALM /ACTS 2/BOOK
DISK(FOIA)/REPORT