Appeal No. 1999-2069
Application 08/397,639
We find, based on Smilansky and Frankot, that persons of
ordinary skill in the image registration art knew that affine
transformations could be used for image registration and that the
method of least squares was used to calculate the factors of
affine transformation. One of ordinary skill in the art also
knew from Kano that a complex nonlinear warping transformation
could be utilized to obtain improved registration between two
images where the distortion is significant. In our opinion, it
would have been obvious to one having ordinary skill in the image
registration art to replace the nonlinear warping transformation
of Kano with any known transformation, including the simpler
affine transformations (rotation, scaling, and translation),
which are taught to be well known in Smilansky and Frankot,
depending on the kind of distortion to be corrected. That is, it
would have been generally obvious to one of ordinary skill in the
art to go from a complex transformation to a known simple
transformation if only a simple transformation is needed. Next,
we consider Appellants' arguments.
Appellants argue (Br5):
Kano et al. is directed to adjusting image interval changes
such as a shape or size of a lung or heart or such as
overlapping portions of ribs or veins. The interval changes
are not analogous shifts, such as enlargement/reduction,
rotation or parallel displacement, which are the subject of
affine transformation. Therefore, Kano et al. is completely
different from the present invention.
- 11 -
Page: Previous 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Next
Last modified: November 3, 2007