FPGA Central

I have a fairly general question for the group. My application is
real time image enhancement using the T.I. C64x. The approach is to
use fast convolution, in the frequency domain, due to the large filter
kernel.

Is real-time image convolution, in the frequency domain, a fairly
common approach? If so, it would seem fixed point
examples/application notes for performing freq domain image
convolution on the popular fixed-point TI DSP would be readily
available. Could anyone point me to any examples/app notes dealing
with getting the image into the freq domain and then back to the
spatial domain?

In the midst of implementation issues, the nagging questions is: am I
reinventing the wheel?

Thanks for any input,

Ted