Backport from master.

- new reduce implementation (with kepler optimizations)
- saturate_cast via asm command
- video SIMD instructions in element operations
- float arithmetics instead of double
- new deviceSupports function

- fixed printCudaDeviceInfo for new CC
- fixed some compilation errors and warnings
- removed unset command from CMake script
- removed unused std imports

used random images in gpu filter tests

added PyrLKOpticalFlowDense performance test

added convert functions to BruteForceMatcher_GPU
other minor fixes

fixed BFM warning ("cannot tell what pointer points to")

minor optimization of SURF_GPU (reduce memory transfers, use structure of arrays instead of array of structures)