Please excuse the following if I've overlooked something obvious, as I am just starting with IPP. I need to calculate the partial derivatives of a 2D image, up to order 2, in both x and y directions. But I'd like to smooth the image at the same time by convolution with Gaussian kernels. Normally this is done by convolving with separable kernels, eg to get the first derivative in the x direction one convolves with a derivative of Gaussian kernel along the rows and a Gaussian kernel along the columns. With the IPP, it seems this could be done with one of the generic convolution functions, but then I'd have to make the kernels myself. This would be OK but I was wondering if there might be a packaged function for the purpose, as its a common operation in image processing.
Thanks,
