Namgoo Lee
4366c8734f
Fix Farneback Optical Flow Algorithm
...
- Before this PR, following tests failed on some platform.
CUDA_OptFlow/FarnebackOpticalFlow.Accuracy/19
CUDA_OptFlow/FarnebackOpticalFlow.Accuracy/23
- The algorithm now recognizes the OPTFLOW_USE_INITIAL_FLOW flag.
Previously, when the flag was set, it did not use the flow data
passed as input, instead used some garbage data in memory.
- More strict test limit.
2019-01-13 15:14:45 +00:00
Hamdi Sahloul
a39e0daacf
Utilize CV_UNUSED macro
2018-09-07 20:33:52 +09:00
dtmoodie
4566105578
fixed non c++11 warnings.
2017-10-15 12:12:00 -04:00
dtmoodie
3c229aa2eb
async bugfix
...
upload winsize and iters as members to prevent async out of scope issues
2017-09-14 14:48:49 +00:00
Tim King
203b398caa
Fixed bug #7482 . Updated dense flow routine to reference bound textures.
2017-01-18 19:30:45 +02:00
Maksim Shabunin
644c0dd5f4
Merge pull request #6940 from andy-held:cuda_pyrlk_static_const_init
2016-07-19 08:47:57 +00:00
Vadim Pisarevsky
d62b0bd363
Merge pull request #6849 from alcinos:optflow_interface
2016-07-18 15:05:13 +00:00
Andreas Franek
13d1ee10be
replaced static const members with static functions for msvc
...
cpmpatability
2016-07-18 14:41:55 +02:00
Andreas Franek
8d43e2b945
fixed the cuda optical flow error normalization factor
...
texture channels were not considered correctly, nor was the cuda texture
normalization
2016-07-04 16:13:31 +02:00
alcinos
e22b838af8
Wrap SparseOptFlow class around PyrLK optical flow computation
2016-01-29 01:47:51 +01:00
alcinos
be4312ec3d
Wrap DenseOptFlow class around Farneback optical flow computation
2016-01-28 20:03:27 +01:00
Dan Moodie
66738d748f
Modified sparse pyrlk optical flow to allow input of an image pyramid which thus allows caching of image pyramids on successive calls.
...
Added unsigned char support for 1, 3, 4 channel images.
2015-12-29 10:48:14 -05:00
Alex Torres
383b7e9dd1
Fixes namespace error on cudaoptflow
2015-07-29 14:22:45 -03:00
Alex Torres
2838ccc78e
Fixes compilation errors when BUILD_CUDA_STUBS is ON
2015-07-29 08:59:38 -03:00
Vladislav Vinogradov
381216aa54
refactor cudaoptflow public API:
...
* use opaque algorithm interfaces
* add stream support
2015-01-21 15:50:21 +03:00
Vladislav Vinogradov
19c6bbe7d9
move obsolete algorithms from cudaoptflow to cudalegacy
2015-01-21 15:50:21 +03:00
Vladislav Vinogradov
71c391cd8d
remove unused memory transfer from TVL1 CUDA implementation
...
it caused runtime failures
2015-01-21 15:50:21 +03:00
Vladislav Vinogradov
9210d8e542
move allocMatFromBuf function to farneback.cpp:
...
* it is the only place, where it is used
* no need to make this function public
2014-12-23 17:42:49 +03:00
Philipp Hasper
30742d21d8
CUDA optical flow stubs: missing parameter
2014-08-25 14:49:21 +02:00
Ernest Galbrun
5623701acb
performance issue for cuda TVL1 when gamma = 0
2014-07-28 14:24:21 +02:00
Ernest Galbrun
62fed8b7b2
retry after failure to load from the build bot
2014-07-07 12:55:07 +02:00
Ernest Galbrun
5c8e679bdc
still a couple tabs and trailing whitespaces...
2014-07-07 12:34:23 +02:00
Ernest Galbrun
eb6c598678
changed nullptr to NULL to avoid c++11 (failed to build on linux)
...
replaces tabs with spaces
2014-07-07 09:32:48 +02:00
Ernest Galbrun
693c4e5741
debug of cuda_tvl1 => pass tests succesfully
2014-07-04 14:23:09 +02:00
Ernest Galbrun
b66a13183e
added cuda support for chambolle parameter
2014-07-02 17:06:52 +02:00
Vladislav Vinogradov
1ad4592bfc
Merge branch 'master' into gpu-cuda-rename
...
Conflicts:
modules/cudaoptflow/perf/perf_optflow.cpp
modules/cudaoptflow/src/tvl1flow.cpp
samples/gpu/stereo_multi.cpp
2013-09-04 09:58:32 +04:00
Vladislav Vinogradov
429bfad225
removed precomp.cpp files
2013-09-02 14:00:44 +04:00
Vladislav Vinogradov
fd88654b45
replaced GPU -> CUDA
2013-09-02 14:00:44 +04:00
Vladislav Vinogradov
5660d6a680
renamed gpuoptflow -> cudaoptflow
2013-09-02 14:00:43 +04:00
