Ex parte GO

Board of Patent Appeals and Interferences

Dec 8, 2000

08425990 (B.P.A.I. Dec. 8, 2000)

The opinion in support of the decision being entered today was not written
for publication in a law journal and is not binding precedent of the Board.
Paper No. 22
UNITED STATES PATENT AND TRADEMARK OFFICE
____________
BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________
Ex parte SHIYU GO
____________
Appeal No. 1998-1895
Application No. 08/425,990
____________
ON BRIEF
____________
Before FLEMING, RUGGIERO, and BARRY, Administrative Patent
Judges.
BARRY, Administrative Patent Judge.
DECISION ON APPEAL
This is a decision on appeal under 35 U.S.C. § 134 from
the rejection of claims 1-6, 8-10, 12-19, 26-33 and 79-86.
We reverse.
BACKGROUND
The invention at issue in this appeal relates to
compressive image encoding and decoding. Compression is
essential to efficient storage and transmission of digitized
Appeal No. 1998-1895 Page 2
Application No. 08/425,990
images. Compression methods have been described by the Joint
Photographic Experts Group (JPEG) for still images and the
Motion Picture Experts Group (MPEG) for moving images. The
JPEG method involves a discrete cosine transform, followed by
quantization and variable-length encoding. The MPEG method
involves detecting motion vectors. Both methods require
extensive computation, with the detection of motion vectors
being particularly demanding.
The appellant's invention encodes a digitized image by
detecting edges in the image, encoding the position and
sharpness of the detected edges, filtering the image by a low-
pass filter to generate a low-frequency image, and encoding
the low-frequency image. A digitized image encoded in this
way is reconstructed by generating a horizontal edge image and
a vertical edge image from the encoded edge position and
sharpness, synthesizing a pair of high-frequency images by
filtering the horizontal and vertical edge images with an edge
synthesis filter, decoding the low-frequency image, and
performing an inverse wavelet transform on the decoded low -
frequency image and the high frequency images. Synthesizing
Appeal No. 1998-1895 Page 3
Application No. 08/425,990
the high-frequency images from the edge images enables high
compression, without recourse to extensive computation.
Claim 1, which is representative for our purposes,
follows:
1. A method of encoding and decoding a
digitized image consisting of pixel values,
comprising the steps of:
detecting sharpness of edges in said digitized
image;
encoding position and sharpness values of edge
points having sharpness values exceeding a certain
threshold, thereby generating edge information;
filtering said digitized image by a low-pass
filter, thereby generating a low-frequency image;
encoding said low-frequency image, thereby
generating low-frequency information;
sending said edge information and said low-
frequency information to an input/output device;
receiving said edge information and said low-
frequency information from said input/output device;
generating a horizontal edge image and a
vertical edge image from said edge information;
synthesizing a pair of high-frequency images by
filtering said horizontal edge image and said
vertical edge image with an edge synthesis filter;
Appeal No. 1998-1895 Page 4
Application No. 08/425,990
decoding said low-frequency information, thereby
obtaining a decoded low-frequency image; and
performing an inverse wavelet transform on said
decoded low-frequency image and said pair of high-
frequency images.
Besides the appellant‘s admitted prior art (AAPA), the
references relied on in rejecting the claims follow:
Schreiber 3,035,121 May 15,
1962
Toriu et al. (Toriu) 4,908,872 Mar. 13, 1990
Ohsawa et al. (Ohsawa) 5,124,811 June 23,
1992
van der Wal 5,359,674 Oct. 25, 1994
filed Dec. 11, 1991
Carnahan 5,414,780 May 9, 1995
filed Jan. 27, 1993.
Claims 1-5, 15-19, 26, and 79 stand rejected under 35 U.S.C.
§ 103 as obvious over Schreiber in view of van der Wal further
in view of Carnahan or AAPA. Claims 6 and 8 stand rejected
under § 103 as obvious over Schreiber in view of van der Wal
further in view of Carnahan or AAPA even further in view of
Ohsawa. Claims 9, 10, 12-14, 27-33, and 80-86 stand rejected
under § 103 as obvious over Schreiber in view of van der Wal
Appeal No. 1998-1895 Page 5
Application No. 08/425,990
further in view of Carnahan or AAPA even further in view of
Toriu. Rather than repeat the arguments of the appellant or
examiner in toto, we refer the reader to the brief and answer
for the respective details thereof.
OPINION
In deciding this appeal, we considered the subject matter
on appeal and the rejection advanced by the examiner.
Furthermore, we duly considered the arguments and evidence of
the appellant and examiner. After considering the totality of
the record, we are persuaded that the examiner erred in
rejecting claims 1-6,
8-10, 12-19, 26-33, and 79-86. Accordingly, we reverse.
We begin by noting the following principles from
In re Rijckaert, 9 F.3d 1531, 1532, 28 USPQ2d 1955, 1956
(Fed. Cir. 1993).
In rejecting claims under 35 U.S.C. Section 103, the
examiner bears the initial burden of presenting a
prima facie case of obviousness. In re Oetiker, 977
Appeal No. 1998-1895 Page 6
Application No. 08/425,990
F.2d 1443, 1445, 24 USPQ2d 1443, 1444 (Fed. Cir.
1992).... "A prima facie case of obviousness is
established when the teachings from the prior art
itself would appear to have suggested the claimed
subject matter to a person of ordinary skill in the
art." In re Bell, 991 F.2d 781, 782, 26 USPQ2d
1529, 1531 (Fed. Cir. 1993) (quoting In re Rinehart,
531 F.2d 1048, 1051, 189 USPQ 143, 147 (CCPA 1976)).
With these principles and finding in mind, we consider the
examiner's rejections and appellant's argument.
The combination of references applied in each of the
examiner's rejections includes the subcombination of Schreiber
in view of van der Wal. Regarding the subcombination, the
examiner asserts, "[i]t would have been obvious ... to use the
pyramid filtering of van der Wal, since the system of van der
Wal also applies to coding as noted in col. 20, lines 63-68,
which is commonly used in subband image coding, because van
der Wal further provides for image enhancement and noise
reduction, and because both Schreiber and van der Wal obtain
edge data." (Examiner's Answer at 5.) The appellant argues,
"van der Wal would not lead an ordinarily skilled person to
modify Schreiber ...." (Appeal Br. at 17.)
Appeal No. 1998-1895 Page 7
Application No. 08/425,990
“Obviousness may not be established using hindsight or in
view of the teachings or suggestions of the inventor.”
Para-Ordnance Mfg., 73 F.3d at 1087, 37 USPQ2d at 1239 (citing
W.L. Gore & Assocs., Inc., 721 F.2d at 1551, 1553, 220 USPQ at
311, 312-13 (Fed. Cir. 1983)). “It is impermissible to use
the claimed invention as an instruction manual or ‘template’
to piece together the teachings of the prior art so that the
claimed invention is rendered obvious.” In re Fritch, 972
F.2d 1260, 1266, 23 USPQ2d 1780, 1784 (Fed. Cir. 1992) (citing
In re Gordon, 733 F.2d 900, 902, 221 USPQ 1125, 1127 (Fed.
Cir. 1984)). "[T]o establish obviousness based on a
combination of the elements disclosed in the prior art, there
must be some motivation, suggestion or teaching of the
desirability of making the specific combination that was made
by the applicant." In re Kotzab,
217 F.3d 1365, 1370, 55 USPQ2d 1313, 1316 (Fed. Cir. 2000)
(citing In re Dance, 160 F.3d 1339, 1343, 48 USPQ2d 1635, 1637
(Fed. Cir. 1998) and In re Gordon, 733 F.2d 900, 902, 221 USPQ
1125, 1127 (Fed. Cir. 1984)).
Appeal No. 1998-1895 Page 8
Application No. 08/425,990
Here, the examiner fails to identify a sufficient
suggestion to combine van der Wal with Schreiber. Schreiber
teaches "bandwidth-reduction systems ...." Col. 1, l. 9. For
its part, van der Wal discloses "multiresolution signal
processing circuitry which has been simplified so that it may
be implemented as a single IC. The circuitry includes a
filter and is configured to accept input signals having
imbedded
timing signals." Col. 2, ll. 50-54. Although van der Wal
teaches that "[i]n this configuration, multiple signal
processing circuits may be coupled in cascade to produce a
multi-stage pyramid processing system", id. at ll. 58-60, the
examiner fails to identify a sufficient suggestion to add the
multi-stage pyramid processing system to the systems of
Schreiber.
As aforementioned, the examiner relies only on certain
lines in van der Wal for a suggestion to combine van der Wal
with Schreiber. To put the lines in context, the full
paragraph of the reference that contains the lines follows.
Appeal No. 1998-1895 Page 9
Application No. 08/425,990
FIG. 10 illustrates a configuration of two
pyramid ICs, 1002 and 1004, and two frame stores
1006 and 1008 as a reduce-expand pyramid module.
This module may be used for encoding and decoding
images for data reduction, image enhancement, noise
reduction image merging and other image processing
functions where it is desirable to reconstruct an
image from a pyramid after some processing is
performed on pyramid images.
Col. 20, l. 63 - col. 21, l. 2. In summary, the paragraph
teaches that van der Wal's reduce-expand pyramid module is
useful "where it is desirable to reconstruct an image from a
pyramid after some processing is performed on pyramid images."
There is no evidence, however, that Schreiber employs, let
alone processes, such pyramid images such that it would be
desirable to reconstruct an image from a pyramid.
Relying on AAPA as evidence that "wavelets provide for
efficiency and high compression ratios" (Examiner's Answer at
6), on Carnahan to "demonstrate[] the commonality of using a
wavelet transform in image coding" (id.), on Ohsawa to
"provide[] for a series of low-pass filters in an encoding
apparatus" and "for different cut-off frequencies" (id. at 8),
and on Toriu "only ... as an example to show the commonality
of a typical horizontal and vertical gradient operations" (id.
Appeal No. 1998-1895 Page 10
Application No. 08/425,990
at 9), the examiner fails to allege, let alone show, that
these references cure the deficiency of Schreiber and van der
Wal. Because there is no evidence that van der Wal's reduce-
expand pyramid module would have been desirable in Schreiber's
systems, we are not persuaded that teachings from the prior
art would have suggested the combination.
In addition, claims 1-6, 8-10, and 12-14 specify in
pertinent part the following limitations: "synthesizing a pair
of high-frequency images by filtering said horizontal edge
image and said vertical edge image with an edge synthesis
filter ... and performing an inverse wavelet transform on said
decoded lowfrequency image and said pair of high-frequency
images." Similarly, claims 26-33 specify in pertinent part
the following limitations: "synthesizing a pair of high-
frequency images by filtering said horizontal edge image
horizontally with an edge synthesis filter, and filtering said
vertical edge image vertically with said edge synthesis
filter; ... performing an inverse wavelet transform on said
high-frequency images and said low-frequency image, thereby
obtaining said digitized image ...." Also similarly, claims
Appeal No. 1998-1895 Page 11
Application No. 08/425,990
79-86 specify in pertinent part the following limitations:
"synthesizing a pair of high-frequency images by filtering
said pair of edge images with an edge synthesis filter; and an
inverse wavelet transform processor, coupled to said up
sampler and said edge synthesizer, for performing an inverse
wavelet transform on said decoded low-frequency image and said
pair of high-frequency images, thereby obtaining said
digitized image." Accordingly, claims 1-6, 8-10, 12-14, 26-
33, and 79-86 require performing an inverse wavelet transform
on a pair of high-frequency images that were synthesized by
filtering a pair of edge images with an edge synthesis filter.
The examiner fails to show a suggestion of the
limitations in the prior art. He admits, "[n]either van der
Wal nor Schreiber explicitly provide for an inverse wavelet
transform on the low and high-frequency images ...."
(Examiner's Answer
at 5.) In terms of the AAPA, the appellant acknowledges "much
interest in the wavelet transform as a means of obtaining high
compression ratios with relatively modest amounts of
computation." (Spec. at 1.) Although the AAPA also discloses
Appeal No. 1998-1895 Page 12
Application No. 08/425,990
the performance of "[a]n inverse wavelet transform ... to
obtain the original image" (id. at 2-3), the inverse transform
is not performed on a pair of high-frequency images that were
synthesized by filtering a pair of edge images with an edge
synthesis filter. To the contrary, the AAPA's inverse wavelet
transform is performed on inter alia high-frequency data
resulting from a wavelet transform. The specific admission
follows.
Another prior-art wavelet encoding scheme
employs a basic wavelet that is the first derivative
of a smoothing filter (that is, the first derivative
of a low-pass filtering function). This type of
wavelet acts as a highpass filter. High-frequency
information is obtained by detecting local peaks
(local maxima of absolute values) in the result of
the wavelet transform, which correspond to edges in
the original image. The size and location of the
peak values at a selected scale are encoded, along
with a low-frequency image obtained by smoothing at
the largest scale of the wavelet transform. Fairly
high compression ratios can be obtained in this way.
To reconstruct the original image from the
encoded data, this prior-art method employs an
algorithm derived from a mathematical procedure
involving iterated projections in Hilbert space.
Under ideal conditions, the projections converge
toward a unique set of data that (i) have the
required local peak values and (ii) are within the
range of the wavelet transform operator. An inverse
wavelet transform is then carried out on the
converged data to obtain the original image.
Appeal No. 1998-1895 Page 13
Application No. 08/425,990
(Id.)
Similarly, although Carnahan discloses an "[i]nverse
image transformation circuit 64", col. 12, l. 67, the circuit
does not perform its inverse transformation on a pair of high-
frequency images that were synthesized by filtering a pair of
edge images with an edge synthesis filter. To the contrary,
the reference's inverse transformation is performed on image
data resulting from a wavelet transform. Specifically,
"[i]nverse image transformation circuit 64 performs
(recursively) the inverse
operations performed by circuit 52. In a preferred
embodiment, circuit 64 has the same structure does [sic]
circuit 52 except that each filter of circuit 64's analyzers
... generates an 'inverse' set of coefficients to the
coefficients generated by the corresponding filter of circuit
52. Each N×M image data block output from circuit 64 is a
reconstructed version of a corresponding N×M image data block
received by circuit 52."
Col. 12, l. 67 - col. 13, l. 8.
Appeal No. 1998-1895 Page 14
Application No. 08/425,990
Relying on Ohsawa and Toriu for the aforementioned
reasons, the examiner fails to allege, let alone show, that
either reference cures the deficiency of Schreiber, van der
Wal, AAPA, and Carnahan. Because AAPA and Carnahan teach
performing an inverse wavelet transform only on data resulting
from a wavelet transform, we are not persuaded that teachings
from the prior art would have suggested the limitations of
"synthesizing a pair of high-frequency images by filtering
said horizontal edge image and said vertical edge image with
an edge synthesis filter ... and performing an inverse wavelet
transform on said decoded lowfrequency image and said pair of
high-frequency images"; "synthesizing a pair of high-frequency
images by filtering said horizontal edge image horizontally
with an edge synthesis filter, and filtering said vertical
edge image vertically with said edge synthesis filter; ...
performing an inverse wavelet transform on said high-frequency
images and said low-frequency image, thereby obtaining said
digitized image"; or "synthesizing a pair of high-frequency
images by filtering said pair of edge images with an edge
synthesis filter; and an inverse wavelet transform processor,
coupled to said up sampler and said edge synthesizer, for
Appeal No. 1998-1895 Page 15
Application No. 08/425,990
performing an inverse wavelet transform on said decoded low-
frequency image and said pair of high-frequency images,
thereby obtaining said digitized image." For the foregoing
reasons, we reverse the rejection of claims 1-5, 15-19, 26,
and 79 as obvious over Schreiber in view of van der Wal
further in view of Carnahan or AAPA; the rejection of claims 6
and 8 as obvious over Schreiber in view of van der Wal further
in view of Carnahan or AAPA even further in view of Ohsawa;
and the rejection of claims 9, 10, 12-14, 27-33, and 80-86 as
obvious over Schreiber in view of van der Wal further in view
of Carnahan or AAPA even further in view of Toriu.

CONCLUSION
In summary, the rejection of claims 1-5, 15-19, 26, and
79 under 35 U.S.C. § 103 as obvious over Schreiber in view of
van der Wal further in view of Carnahan or AAPA is reversed.
The rejection of claims 6 and 8 under § 103 as obvious over
Schreiber in view of van der Wal further in view of Carnahan
or AAPA even further in view of Ohsawa is also reversed. In
addition, the rejection of claims 9, 10, 12-14, 27-33, and 80-
86 under § 103 as obvious over Schreiber in view of van der
Appeal No. 1998-1895 Page 16
Application No. 08/425,990
Wal further in view of Carnahan or AAPA even further in view
of Toriu is reversed.
Appeal No. 1998-1895 Page 17
Application No. 08/425,990
REVERSED
MICHAEL R. FLEMING )
Administrative Patent Judge )
)
)
)
) BOARD OF PATENT
JOSEPH F. RUGGIERO ) APPEALS
Administrative Patent Judge ) AND
) INTERFERENCES
)
)
)
LANCE LEONARD BARRY )
Administrative Patent Judge )
Appeal No. 1998-1895 Page 18
Application No. 08/425,990
VENABLE
POST OFFICE BOX 34385
WASHINGTON, DC 20043-9998
LLB/dal