I am using three pairs of separable filters.
Each pair has a X-kernel and a Y-Kernel.
each kernel has 15 floating point values.
I am using the H.264 modules from the IPP samples.
I am filtering each frame (YUV) with these three pairs of kernels one after the other, which turns out to be quite time consuming.
Is it possible to derive one filter which can do the filtering done by the three filters?
Please feel free to share you thoughts.