CUDA H.264 decoding to make my own post-processing.

Hi,

I am looking for and example code to understand how to decode H.264 using the GPU. My goal is to receive some H.264 video stream from an IP camera and obtain RGB frames in order to make my own kernels to post-process each image.

Let´s see if some nVidia engineer or any other person has some easy undestanding code for a CUDA starter. It is just to see in a secuantial main how to use de videodecoder and videparser that i have seen in nvidia examples. I would like, if it is posible, to find the way to discard frames that are not diferent enough so I dont post-process them.

Many thanks friends!

PD: Sorry for my english, i am an spanish starter developer.