How to input Image to convolution separable program

I am not able to write the code. Is it same as bmploader.cpp in image denoising program which is given in sdk. if its different please tell me the simpler one.

Thanks for spending time on this.