OpenCV 3.0 prebuilt with CUDA using Toolkit 7.0 please
Hello! Can somebody do a reliable build of openCV 3.0 prebuilt with CUDA toolkit 7.0? It could go on the source forge page with the rest of the prebuilds.
I've recently got a laptop with a nVidia GTX 960m (Maxwell architecture) and the current openCV 3.0 beta build with CUDA 6.5 doesn't run cudaimgproc.hpp processes, such as cvtColor for changing color spaces, (a few other things don't work either such as a simple gpu_Mat.cols) the program just hangs with no warnings or errors as soon as I use a function that comes from cudaimgproc.hpp. The same build works fine my my desktop nVidia Quadro 4000. I have both CUDA toolkit 6.5 and 7.0 installed on my laptop, as well as Microsoft Visual Studio 2013 on Windows 8.1
Cmake seems to just create a lot of problems, it would be much easier if somebody who knows what they are doing could build one for the rest of us with no idea about dealing with cmake errors.
Many thanks and you'll be doing a great service to new users to openCV!