It’s one thing to compare GPU code performance with CPU code performance. If the algorithm is parallel, GPU would beat CPU any day. In our case, CUVI beats the best (performance wise) CPU primitives library on the planet, Intel(r) IPP. Take a look at the performance figures.
But this is not what this blog post is about. Willow Garage recently released new version of OpenCV with added GPU functionality. Naturally, we wanted to give it a test drive. Surprisingly, in some high value, compute intensive imaging functions, CUVI won over OpenCV GPU! We chose Optical Flow Lucas & Kanade (Pyramid) and compared the performance of CUVI’s version with OpenCV’s version. Here are the results:
You can try it out yourself with CUVI v1.2 now available for download.