Vladislav Vinogradov 281d036fcf optimizations: ...

- 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

2013-01-23 21:05:07 +04:00

..

opencv2/gpu

optimizations:

2013-01-23 21:05:07 +04:00