Hello, I'm on about my second week attempting to port a very very large opencv project to IPP. I ported most of the functionality to OpenCV allowing it to exist independantly of IPL/IPP differences. I had a few questions that maybe someone would be able ot help me with.
First, we were using iplConvolveSep2DFP for which I am not sure if I can use the x/y separate kernels. The best I can guess would be to filter it twice maybe using two different kernels?
Second, other than cvGet2D, which I have had a lot, and I mean a lot of trouble with, is there a way to address a single pixel in the image? we've been calculating it with ptr +width*y + x which doesn't always work and is ugly and nasty. We noticed that in the ipp-ipl-migration function mapping that it said ippiCopy which isn't obvious how to copy a single pixel. Could this be clarified?
and also, I've seemed to have lost this file, is it lost somewhere in the directories or on the net?
Thanks a bunch!
~Nick
