Ex parte Baumgart et al.

Board of Patent Appeals and Interferences

Sep 15, 1998

08039674 (B.P.A.I. Sep. 15, 1998)

Application for patent filed March 30, 1993. 1
1
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publication in a law journal and (2) is not binding precedent of the Board.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. 11
UNITED STATES PATENT AND TRADEMARK OFFICE
____________
BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________
Ex parte Chris W. Baumgart and Christopher A. Ciarcia
____________
Appeal No. 95-1217
Application 08/039,6741
____________
Before HAIRSTON, BARRETT, and FLEMING, Administrative Patent
Judges.
FLEMING, Administrative Patent Judge.
DECISION ON APPEAL
This is a decision on appeal from the final rejection of
claims 1 through 13 and claims 15 through 20, all of the
claims pending in this application. Claim 14 has been
canceled.
The invention is directed to an automatic target
recognition apparatus and method. The invention can be used
Appeal No. 95-1217
Application 08/039,674
2
to recognize an object, such as a land mine, within a
digitized image. In the
claimed invention, a camera/digitizer transmits a digitized
image signal to a computer. The computer processes the image
by using a number of different analysis chains. Examples of
analysis techniques that may be used in these analysis chains
are object texture analysis, background subtraction, and
object edge enhancement. The analysis chains are shown in
Figure 3. Each of the analysis chains receives the image
signal and analyzes it in parallel with the other analysis
chains. The information obtained in each of the analysis
chains can be combined to obtain a single result.
Independent claim 1 is reproduced as follows:
1. An automatic target recognition apparatus for
recognizing an object within a digitized image, comprising:
a video camera and digitizer for producing a
digitized image;
and a computer for processing the digitized image,
wherein;
the image is processed in a plurality of parallel
analysis chains, each of said analysis chains being a distinct
means for analyzing the image such that the object may be
identified by one or more of said parallel analysis chains.
Appeal No. 95-1217
Application 08/039,674
3
The Examiner relied on the following references:
Corwin et al. 5,233,541 8/93
Fukumizu 5,060,278 10/91
Crimmins et al. 4,644,585 2/87
Prakash 5,054,101 10/91
Huynh et al. 4,878,114 10/89
Eckstein, Jr. 3,947,833 3/76
Natakani 4,817,174 3/89
Hunt et al. 4,335,427 6/82
Claims 1 through 13 and claims 15 through 20 stand
rejected under 35 U.S.C. § 112, ¶ 1, as based on a non-
enabling disclosure. Claims 3, 7 through 11, and 17 through
18 stand rejected under 35 U.S.C. § 112, ¶ 2, as indefinite
due to failure to point out and distinctly claim the
invention.
The claims were also rejected under 35 U.S.C. § 103.
Claims 1, 5, 12, 13, and 15 stand rejected under § 103 as
unpatentable over Corwin et al. in view of Fukumizu. Claims
2, 7, 8, 11, 14, and 18 stand rejected under § 103 as
Appeal No. 95-1217
Application 08/039,674
4
unpatentable over Corwin et al. in view of Fukumizu and
further in view of Crimmins et al. Claims 3 and 17 stand
rejected under § 103 as unpatentable over Corwin et al. in
view of Fukumizu and further in view of Prakash. Claims 4 and
16 stand rejected under § 103 as unpatentable over Corwin et
al. in view of Fukumizu and further in view of Huynh.
Claim 6 stands rejected under § 103 as unpatentable over
Corwin et al. in view of Fukumizu and further in view of
Eckstein, Jr. Claim 9 stands rejected under § 103 as
unpatentable over Corwin et al. in view of Fukumizu and
Crimmins et al. and further in view of Hunt et al. Claim 10
stands rejected under § 103 as unpatentable over Corwin et al.
in view of Fukumizu and Crimmins et al. and further in view of
Huynh et al. Claim 19 stands rejected under § 103 as
unpatentable over Corwin et al. in view of Fukumizu and
further in view of Nakatani. Finally, claim 20 stands
rejected under § 103 as unpatentable over Corwin et al. in
view of Fukumizu and further in view of Hunt et al.
An amendment was filed after the March 21, 1994, Office
Action (final). This amendment was not entered.
Appeal No. 95-1217
Application 08/039,674
5
Rather than reiterate the arguments of Appellants and the
Examiner, reference is made to the brief and answer for the
respective details thereof.
OPINION
We will not sustain the rejection of any of the claims.
Specifically, we reverse the rejection of claims 1 through 13
and
claims 15 through 20 under 35 U.S.C. § 112, ¶ 1; we reverse
the
rejection of claims 3, 7 through 11, and 17 through 18 under
35 U.S.C. § 112, ¶ 2; and we reverse the rejection of claims 1
through 13 and claims 15 through 20 under 35 U.S.C. § 103.
1. The rejections under § 112, ¶ 1
In order to be enabling under 35 U.S.C. § 112, a patent
application must sufficiently disclose an invention to enable
those skilled in the art to make and use it. In re Buchner,
929 F.2d 660, 661, 18 USPQ2d 1331, 1332 (Fed. Cir. 1991).
"Although not explicitly stated in section 112, to be
Appeal No. 95-1217
Application 08/039,674
6
enabling, the specification of a patent must teach those
skilled in the art how to make and use the full scope of the
claimed invention without 'undue experimentation.'" In re
Wright, 999 F.2d 1557, 1561, 27 USPQ2d 1510, 1513 (Fed. Cir.
1993) citing In re Vaeck, 947 F.2d 488, 495, 20 USPQ2d 1438,
1444 (Fed. Cir. 1991); In re Wands, 858 F.2d 731, 736-37, 8
USPQ2d 1400, 1404 (Fed. Cir. 1988); In re Fisher, 427 F.2d
833, 839, 166 USPQ 18, 24 (CCPA 1970). "When rejecting a
claim under the enablement requirement of section 112, the PTO
bears an initial burden of setting forth a
reasonable explanation as to why it believes that the scope of
protection provided by that claim is not adequately enabled by
the description of the invention provided in the specification
of the application; this includes, of course, providing
sufficient reasons for doubting any assertions in the
specification as to the scope of enablement." In re Wright,
999 F.2d at 1561-62, 27 USPQ2d at 1513.
Appeal No. 95-1217
Application 08/039,674
7
The Examiner provided a number of reasons for rejecting
claims 1 through 13 and claims 15 through 20 as non-enabling.
The first reason was that the drawings are missing several
elements that are referred to in the specification. Page 8 of
the specification refers to a background "37" of the image,
but item 37 is not shown in Figure 2. The specification
defines the background of the image as the portion which does
not contain objects. The Examiner has failed to provide any
explanation as to why a person of ordinary skill in the art
would be unable to make and use the claimed invention due to
the Appellants’ failure to label background 37 in Figure 2.
The rejection is therefore reversed.
Similarly, the specification refers to edge trace
operations (57, 73, 81, 99, 113, and 137) as steps in certain
of the
analysis chains that are described. It also refers to "second
shadow subtract operation 127" as a step in one of the
analysis chains that is described. These items are not
labeled in the block diagram of the image analysis subroutine
shown in Figure 3. The Examiner has failed to provide any
Appeal No. 95-1217
Application 08/039,674
8
explanation, however, as to why a person of ordinary skill in
the art would be unable to make and use the claimed invention
due to the Appellants’ failure to label these items in Figure
3. To the contrary, given that the specification describes
the steps for each analysis chain in sequential order, a
person of ordinary skill in the art would understand where the
missing items should be located in Figure 3. The rejection is
therefore reversed.
The Examiner argues that the specification does not
clearly show how the output of step 90 is combined with the
output of step 92 and how the output of step 124 is combined
with the output of step 126. The Examiner has failed to
provide an adequate explanation as to why it would require
undue experimentation for a person of ordinary skill in the
art to determine how to combine the outputs of these steps in
order to make and use the claimed invention. The rejection is
therefore reversed.
The Examiner argues that the specification should not use
the reference number "35" after the word "pixel" each time
that the word is encountered. In the Examiner’s view, using
Appeal No. 95-1217
Application 08/039,674
9
the reference number in this way is unnecessary and could
cause confusion because different entities such as "three by
three pattern of pixels" and "center pixel" are followed by
the same reference number "35". While it might be clearer if
the reference number "35" were not used in this way, the
Examiner has not shown that a person of ordinary skill in the
art would be unable to make and use the claimed invention due
to this repeated use of the reference number "35". The
rejection is therefore reversed.
Finally, the Examiner argues that the constant "C", which
is contained in a formula given on page 16 of the
specification, cannot be inferred from the context of the
discussion. According to the specification, the constant C is
used to scale the product of E f e to within the range zeron n xn
through ten. To determine the value for this constant, a
person of ordinary skill in the art would simply need to know
the minimum and maximum possible values for the product of En
f e . The Examiner has failed to n xn
Appeal No. 95-1217
Application 08/039,674
10
provide an adequate explanation as to why it would require
undue
experimentation for a person of ordinary skill in the art to
determine these values. Further, the Examiner has failed to
explain why a person of ordinary skill in the art would be
unable to make and use the claimed invention without knowing
the constant C. The rejection is therefore reversed.
2. The rejections under 35 U.S.C. § 112, ¶ 2
Analysis of 35 U.S.C. § 112, second paragraph, should
begin with the determination of whether the claims set out and
circumscribe the particular area with a reasonable degree of
precision and particularity; it is here where definiteness of
the language must be analyzed, not in a vacuum, but always in
light of teachings of the disclosure as it would be
interpreted by one possessing ordinary skill in the art. In
re Johnson, 558 F.2d 1008, 1015, 194 USPQ 187, 193 (CCPA
1977), citing In re Moore, 439 F. 2d 1232, 1235, 169 USPQ 236,
238 (1971). "The test for definiteness is whether one skilled
in the art would understand the bounds of the claim when read
Appeal No. 95-1217
Application 08/039,674
11
in light of the specification." Miles Labs. Inc. v. Shandon
Inc., 997 F.2d 870, 875, 27 USPQ2d
1123, 1126 (Fed. Cir. 1993) (citing Orthokinetics, Inc. v.
Safety Travel Chairs, Inc., 806 F.2d 1565, 1576 (Fed. Cir.
1986)).
The Examiner rejected claim 3 as indefinite because it
provides that the fractal dimension is the "ratio of the
perimeter of the object relative to the surface area of the
object" but does not recite actually determining the ratio of
the perimeter relative to the surface area. Claim 3 is
reproduced as follows, with emphasis added to the relevant
language:
3. The target recognition apparatus of claim 1, wherein:
a fractal dimension of objects within the digitized
image is obtained in one or more of said chains, the fractal
dimension being a ratio of the perimeter of the object
relative to the surface area of the object.
The Examiner has not explained why a person of ordinary
skill in the art would be unable to understand the bounds of
Appeal No. 95-1217
Application 08/039,674
12
claim 3. A person of ordinary skill in the art would
understand that claim 3 requires the determination of the
ratio even though the determination of the ratio is not
separately recited. The rejection is therefore reversed.
The Examiner rejected claims 7 through 11 on the grounds
that the phrases such as "relatively dark", "relatively
light",
and "lighter" are vague and indefinite as used in the claims.
In the Examiner’s view, it is not clear what the criteria are
Appeal No. 95-1217
Application 08/039,674
13
for determining the relativity. Representative claim 7 is
reproduced as follows, with emphasis added to the relevant
language:
7. A method for using a computer to recognize objects
within a digitized video image, the method comprising:
processing the digitized video image in a plurality
of parallel processing chains, wherein;
one or more of the processing chains for analyzing
the content of the video image includes a series of processing
steps for identifying a relatively light target against a
relatively dark background, and;
one or more of the processing chains includes a
numerical inversion operation for inverting the shading within
the image such that a relatively dark image on a relatively
light background becomes equivalent to a relatively light
image on a relatively dark background.
This rejection is reversed because a person of ordinary
skill in the art would understand the bounds of the claim.
They would understand, for example, that because it recites a
"relatively light target against a relatively dark
background", claim 7 requires that the target must be lighter
than the background. The lightness of the target and darkness
of the background are viewed in relation to each other.
Appeal No. 95-1217
Application 08/039,674
14
The Examiner rejected claim 17 on the grounds that the
claim language that recites the "ratio of the perimeter of
each of the objects to the area of each such object" is vague
and indefinite. According to the Examiner, it is not clear
which of the objects are indicated. Claim 17 is reproduced as
follows, with emphasis added to the relevant language:
17. The computerized mine detection apparatus of claim
12, wherein:
one or more of the processing series determines a
fractal dimension of the objects, the fractal dimension being
the ratio of the perimeter of each of the objects to the area
of each such object.
Again, this rejection is reversed because a person of
ordinary skill in the art would understand the bounds of the
claim. They would understand that the objects referred to in
claim 17 are the objects that are being sensed in claim 12.
They would also understand that, in claim 17, the fractal
dimension of the object is determined by dividing the
perimeter of each object by the area of that object.
Finally, the Examiner rejected claim 18 under § 112, ¶ 2,
on the grounds that the term "light intensity" lacks
antecedent
Appeal No. 95-1217
Application 08/039,674
15
basis and is meaningless in the context of the claim. In the
Examiner’s view, it is not clear what role "light intensity"
plays in a digitized image. The Examiner states that it is
also not clear how light intensity is inverted. Claim 18 is
reproduced as follows, with emphasis added to the relevant
language:
18. The computerized mine detection apparatus of claim
12, wherein:
one or more of the processing series inverts a light
intensity of each of portion of the object within the image
prior to discriminating the objects.
This rejection is reversed because a person of ordinary
skill in the art would understand the bounds of the claim. It
is common to refer to a data element in a digital environment
by the names of the real world information which the data
element represents. Thus, a person of ordinary skill in the
art would understand that the phrase "light intensity" refers
to a numerical value associated with a portion of the image
based on a determination as to the intensity of the light in
that portion of the image. They would further understand that
claim 18 calls for the inversion of this numerical value.
Appeal No. 95-1217
Application 08/039,674
16
Claim 18 is not indefinite under § 112, ¶ 2, even though the
phrase "light
intensity" was not introduced earlier in the claim, because a
person of ordinary skill in the art would still understand the
bounds of the claim.
Although the Examiner did not point it out, claim 18
appears to contain a typographical error. It appears to
contain an extra "of" after the word "each". If so, the
phrase "of each of portion" should be changed to "of each
portion". Even with this
typographical error, a person of ordinary skill in the art
would still understand the bounds of the claim.
3. The rejections under 35 U.S.C. § 103
We will not sustain the rejection of claims 1 through 13
and claims 15 through 20 under 35 U.S.C. § 103. It is the
burden of the Examiner to establish why one having ordinary
skill in the art would have been led to the claimed invention
by the express teachings or suggestions found in the prior
art, or by implications contained in such teachings or
suggestions. In re Sernaker, 702 F.2d 989, 995, 217 USPQ 1, 6
Appeal No. 95-1217
Application 08/039,674
17
(Fed. Cir. 1983). The Examiner has failed to set forth such a
prima facie case.
a. Claims 1 through 6
The rejection of independent claim 1, and dependent
claims 2 through 6, is based upon the Fukumizu reference
teaching a
plurality of analysis chains that are each, in the language of
claim 1 (with emphasis added), "a distinct means for analyzing
the image such that the object may be identified by one or
more of said parallel analysis chains." We reverse these
rejections because Fukumizu does not teach this element of
claims 1 through 6.
Fukumizu discloses a pattern recognition apparatus that
uses a neural network system. Pattern data is input into the
system
via a pattern input means, such as a scanner. Feature data is
then extracted from the pattern data and input into a
plurality of neural networks NET , with "i" representing thei
number of neural networks. See col. 3, lines 41-49. Each of
the neural networks corresponds to a certain identification
Appeal No. 95-1217
Application 08/039,674
18
class, each class being a known pattern that the system is
looking to identify. See col. 3, lines 33-35, 47-49. The
"neural networks operate in parallel and as a whole configure
a single neural network system." See col. 3, lines 49-51.
Each of the neural networks NET judges whether the inputi
feature vector belongs to the class C , see col. 3, lines 62-i
65, and outputs the probability that the feature vector
belongs to that class.
See col. 4, lines 4-8. The judgment unit 14 then judges which
of the classes is likely to be the correct class, based on the
probability data output from the neural networks, and outputs
a result. See col. 4, lines 11-21.
The rejection of claims 1 through 6 turns on how the term
"distinct" in claim 1 is interpreted. During patent
prosecution, claims must be given their broadest reasonable
interpretation. See In re Morris, 127 F.3d 1048, 1054-55, 44
USPQ2d 1023, 1027-29
(Fed. Cir. 1997). The first place to look when interpreting a
patent claim is the words of the claim themselves. See
Appeal No. 95-1217
Application 08/039,674
19
Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582, 39
ordinary meaning unless the specification or file history
clearly state that a special definition is intended. Id.
The Appellants’ specification does not define the term
"distinct." In ordinary usage, the term "distinct" can mean
"separate" or it can mean "not alike." See Webster’s New
Universal Unabridged Dictionary 534 (deluxe 2d. ed., Dorset &
Baber 1983). Each of the neural networks (i.e., analysis
chains) in Fukumizu are separate from the other neural
networks. If
"distinct" could be interpreted as meaning "separate", then
Fukumizu would disclose this limitation of claim 1. However,
when the term "distinct" in claim 1 is read in light of the
other language of the claim, and in light of the
specification, it must be interpreted as requiring that each
analysis chains is not like the other analysis chains.
Appeal No. 95-1217
Application 08/039,674
20
Claim 1 recites, in relevant part and with emphasis
added, that "the image is processed in a plurality of parallel
analysis
chains, each of said analysis chains being a distinct means
for analyzing the image." The recitation of a "plurality" of
analysis chains itself requires multiple, separate analysis
chains. The requirement that the analysis chains be
"distinct" would be redundant with the "plurality" requirement
if "distinct" meant that the analysis chains need to be
separate from one another. Because each of the words in a
claim must be assumed to have some meaning, it would not be
reasonable to interpret "distinct" as synonymous with
"separate." Thus, the term "distinct" must require that the
analysis chains are not alike.
The specification supports this interpretation of claim
1. The specification states on page 8 that the preferred
embodiment includes the following analysis chains: a positive
texture
chain, a negative texture chain, a positive background
subtract chain, a negative background subtract chain, a
positive edge enhancement chain, and a negative edge
Appeal No. 95-1217
Application 08/039,674
21
enhancement chain. The bulk of the specification (pages 8-12)
describes how each of these different analysis chains
operates. Because the specification clearly discloses
analysis chains that are not
alike, it supports interpreting the "distinct" limitation in
claim 1 as requiring analysis chains that are not alike.
In view of the above, multiple copies of the same
analysis chain are not distinct means for analysis within the
meaning of claim 1. To be distinct means for analysis,
different calculation processes must be done. If the same
calculation is done in parallel, these parallel processes do
not constitute distinct analysis chains.
The neural networks in Fukumizu all do the same
calculation. Each neural network has been taught to recognize
a different class and therefore each has different weight data
which it uses in performing the calculation. See col. 7, line
65 to col. 8, line 22. Thus, the neural networks do not
constitute "distinct" means for analyzing within the meaning
of claim 1.
Appeal No. 95-1217
Application 08/039,674
22
b. Claims 7 through 11
The rejection of claim 7 is based upon the teachings of
the Corwin et al. and Fukumizu, as applied to claim 1, taken
in view of Crimmins et al. We reverse the rejection of claim
7, and of dependent claims 8 through 13, because the Examiner
has not shown that there is a motivation to combine Crimmins
et al. with Corwin
et al. and Fukumizu. See In re Rouffet, 47 USPQ2d 1453, 1456
(Fed. Cir. 1998) (holding that, when a rejection is based on a
combination of prior art references, there must be some
teaching, motivation, or suggestion to combine the
references).
Claim 7 recites that the digitized video image is
processed "in a plurality of parallel processing chains" and
that "one or more of the processing chains includes a
numerical inversion operation for inverting the shading within
the image such that a relatively dark image on a relatively
light background becomes equivalent to a relatively light
image on a relatively dark background." The Examiner stated
that Crimmins et al. discloses the concept of numerical
inversion that is recited in claim 7. In the Examiner’s view,
Appeal No. 95-1217
Application 08/039,674
23
the motivation for combining the numerical inversion of
Crimmins et al. with Corwin’s automatic
target recognition scheme and Fukumizu’s neural network system
is offered in Crimmins et al., which states that "the
extension of machine vision to industrial or military
operations requiring the detection of a more general classes
of shapes and/or patterns has met with limited results." See
Crimmins et al., col. 1, lines 24-26.
Crimmins et al. discloses a method and apparatus for
automatic shape recognition in an image that is represented by
a matrix of digital data signals. The method disclosed in
Crimmins et al. includes the steps of:
"(a) computing the complement of the first image matrix;
(b) creating a first structuring element,
representative of the shape to be recognized, as a
matrix of digital data signals slightly larger in
dimension than the shape;
(c) creating a second structuring element equal to
the window complement of the first structuring
element;
(d) eroding the first image matrix with the first
structuring element to form a first transformation
matrix;
Appeal No. 95-1217
Application 08/039,674
24
(e) eroding the complement image matrix with the
second structuring element to form a second
transformation matrix; and
(g) combining the corresponding points of the
first and second transformation matrices to form a
result matrix wherein each non-zero point identifies
an origin point where the shape has been recognized
in the first image matrix."

See col. 2, lines 15-31 (indentation added). Nothing about
this method suggests that it would be advantageous to use any
of these steps in a neural network system. Neural networks
operate in an entirely different manner than the Crimmins et
al. method. The
statement in Crimmins et al. which explains that prior art
machine vision systems had met with limited results does not
provide the motivation for doing image inversion in a neural
network system. Because the Examiner has not provided a
sufficient motivation to combine these references, we reverse
the rejection of claims 7 through 11.
c. Claims 12 through 20
The rejection of claims 12, and dependent claims 13
through 20, is based upon the Fukumizu reference teaching a
Appeal No. 95-1217
Application 08/039,674
25
plurality of processing series that can, in the language of
claim 12, "independently processing the video image such that
each of said processing series can independently determine if
a mine is present in the image." We reverse the rejection of
these claims because the neural networks in Fukumizu are not
capable of independently processing an image to determine if a
mine is present in the image. In Fukumizu, each of the neural
networks
determines whether the input vector belongs to a certain
identification class, each class being a known pattern that
the system is looking to identify. See col. 3, lines 33-35,
47-49,
62-65. The neural networks do not operate independently to
identify the object, but rather the "neural networks operate
in parallel and as a whole configure a single neural network
system." See col. 3, lines 49-51.
Appeal No. 95-1217
Application 08/039,674
26
We have not sustained the rejection of claims 1 through
13 and claims 15 through 20 under 35 U.S.C. § 112, ¶ 1; of
claims 3, 7 through 11, and 17 through 18 under 35 U.S.C.
§ 112, ¶ 2; or of claims 1 through 13 and claims 15 through 20
under 35 U.S.C. § 103 claims. Accordingly, the Examiner's
decision is reversed.
REVERSED
KENNETH W. HAIRSTON )
Administrative Patent Judge )
)
)
) BOARD OF PATENT
LEE BARRETT ) APPEALS AND
Administrative Patent Judge ) INTERFERENCES
)
)
)
MICHAEL R. FLEMING )
Administrative Patent Judge )
Richard E. Constant
Assistant General Counsel for Patents
Mail Code GC-42 (FORSTL) MS-6F-067
1000 Independence Avenue, SW., US DOE
Washington, DC 20585