Appeal No. 1999-2069
Application 08/397,639
that a less complex transformation, such as the affine
transformations of Smilansky, could be used if the image
registration did not need to be as accurate or address the same
kind of distortions. It would have been obvious to one of
ordinary skill in the art to replace a complex transformation
with a known simple transformation, where one skilled in the art
would have appreciated the limitations in image registration from
making such a substitution.
Appellants argue that while Frankot describes affine
transformations as an example of a means for adjusting positions
using a selected subarea, "Frankot et al. does not teach that
shifts in position among images are the problem when adding
and/or subtracting radiation images, and also does not teach that
affine transformation is used to solve the problem" (Br6).
Kano, not Frankot, is relied on for teaching subtracting
registered radiation images. The general problem faced by
Frankot and Kano is image registration. Frankot discloses affine
transformation for image registration.
Appellants argue that one of ordinary skill would not have
replaced the nonlinear warping of Kano with the affine
transformation in Frankot (Br6). The Examiner points out that
Frankot mentions second and higher order transformation models in
addition to the first order affine transformation model and that
- 14 -
Page: Previous 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Next
Last modified: November 3, 2007