Appeal No. 2006-1618
Application No. 10/046,797
rather determined from the approximation technique performed by polygonal
approximation block 100 [Kim, col. 3, lines 30-34].
Turning to the secondary reference, Suzuki discloses detecting the
contour of an object within a sequence of images where the user specifies points
adjacent boundary locations of the object in a first image. Contours of
subsequent images are detected based on contour points detected by the first
image [Suzuki, abstract, col. 2, lines 1-50]. Based on the record before us, we
find that the teachings of Suzuki are reasonably combinable with Kim essentially
for the reasons stated by the examiner. Although manually selecting points
would arguably slow Kim's encoding process, we disagree with appellant that
such manual selection would inevitably frustrate Kim's purpose of encoding
video. On the contrary, we see no reason why such manual selection would not
be useful and beneficial to Kim's system -- namely to enable the user to precisely
dictate and control the entry of vertices corresponding to a particular object,
particularly with complex contours. Moreover, we agree with the examiner that
applying the teachings of Suzuki would not slow Kim's system unacceptably
since Suzuki does not require user input in every frame of video; rather, only one
user input is required. After such user input, contours are then detected
automatically in subsequent frames as the examiner indicates.
Furthermore, even assuming that the starting vertices in Kim are
determined automatically, manually selecting starting vertices would add
numerous potential candidates for starting vertices for detecting contours of a
8
Page: Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Next
Last modified: November 3, 2007