Appeal No. 2006-1618
Application No. 10/046,797
using data representing grey levels of the image of the amplitudes and/or
orientations of the gradients as starting data to elaborate image contours
[answer, page 9]. The examiner then finds that it would have been obvious to
one of ordinary skill in the art at the time of the invention to use gradient values
for contour detection in Kim and Suzuki to better represent contours [answer,
page 10]. Regarding claim 4, the examiner further notes that Catros also uses
amplitude values as weights and reduces the contour detection problem to a
"shortest path" problem.
Appellant responds that no motivation exists to combine the references
[brief, page 14]. Specifically, appellant argues that the skilled artisan concerned
with video encoding of known complete boundary image data in Kim would not
be motivated to look to disparate teachings of Catros regarding bridging
disjointed contour elements [id.]. The examiner responds by noting, among other
things, that if Kim's invention were used on an image that contained missing
contours, the skilled artisan would reasonably refer to the teachings of Catros for
a solution to fill in the missing contours [answer, pages 10 and 11]. Appellant
responds that combining Catros with Kim and Suzuki in the manner suggested
by the examiner is improper since, among other things, neither Kim nor Suzuki is
concerned with such discontinuities [reply brief, page 7].
We will sustain the examiner's rejection of claims 3 and 4. We find that
the teachings of Catros are reasonably combinable with Kim and Suzuki
essentially for the reasons stated by the examiner. Indeed, any two adjacent
18
Page: Previous 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Next
Last modified: November 3, 2007