Vladislav Vinogradov 4382302a6b fix matrix type for keypoints buffer in CUDA FAST ...

use CV_32FC1 instead of CV_16SC2 since detectAsync uses CV_32FC1
to reallocate the matrix

2016-07-06 14:27:08 +03:00

..

cuda

Thrust allocator usage.

2016-03-15 11:14:47 -04:00

brute_force_matcher.cpp

…

fast.cpp

fix matrix type for keypoints buffer in CUDA FAST

2016-07-06 14:27:08 +03:00

feature2d_async.cpp

…

orb.cpp

Fix issue #6473

2016-05-20 13:19:14 +03:00

precomp.hpp

…