marina.kolpakova
|
0bf10c9a47
|
added connectivityMask calculation function
|
13 years ago |
|
marina.kolpakova
|
a9764dd910
|
fixed not defined macro warningds
|
13 years ago |
|
marina.kolpakova
|
51433a9331
|
C.C. Labeling: added stream support
|
13 years ago |
|
marina.kolpakova
|
350621057f
|
connected components labeling
|
13 years ago |
|
marina.kolpakova
|
8d73bbb8b7
|
fixed 2228
|
13 years ago |
Vladislav Vinogradov
|
b43cec3301
|
added gpu::graphcut for float sources (CUDA 5.0)
|
13 years ago |
|
Vladislav Vinogradov
|
4f99f69a29
|
added gpu alpha premultiplication
|
13 years ago |
|
Vladislav Vinogradov
|
79d0dc25f4
|
added gpu RGB<->Lab and BGR<->Luv conversions
|
13 years ago |
|
Vladislav Vinogradov
|
051adcb786
|
added gpu BGR<->Lab and RGB<->Luv color conversion and gammaCorrection
|
13 years ago |
Anatoly Baksheev
|
2b0c0e71bb
|
test
|
13 years ago |
|
marina.kolpakova
|
e1e0c46639
|
integrated patch: HOG confidence calculation. Thanks, Wongun.
|
13 years ago |
|
marina.kolpakova
|
4fa282e1a9
|
fixed SURf according to clockwise descriptor convention
|
13 years ago |
|
Anatoly Baksheev
|
1efb3c8235
|
test
|
13 years ago |
|
marina.kolpakova
|
1b7ad93dc9
|
LBP features integrated in CascadeClassifier_GPU
|
13 years ago |
|
marina.kolpakova
|
2dc93574e1
|
LBP: switched to PImpl pattern
|
13 years ago |
|
marina.kolpakova
|
6c3eeb7d7c
|
removed division
|
13 years ago |
|
marina.kolpakova
|
8fb1f4093b
|
removed unused code
|
13 years ago |
|
Marina Kolpakova
|
e63ab8dec5
|
LBP: multiscale approach; refactored atomics usage
|
13 years ago |
|
Marina Kolpakova
|
336dcbcff7
|
fixed brocken tests by classifier loger from r9027
|
13 years ago |
|
Anatoly Baksheev
|
a2430afcac
|
minor LBP for GPU
|
13 years ago |
|
Marina Kolpakova
|
b9e53ec8ea
|
fixed build after r9027
|
13 years ago |
|
Anatoly Baksheev
|
bfe6e2c4b1
|
minor in LBP for GPU
|
13 years ago |
|
Anatoly Baksheev
|
9706079ace
|
a bit refactoring in LBP face detection on GPU
|
13 years ago |
|
Marina Kolpakova
|
58c67ccfea
|
improved LBP indexing
|
13 years ago |
|
Marina Kolpakova
|
965109228d
|
added delobal memory version
|
13 years ago |
|
Marina Kolpakova
|
2aacff4c39
|
swithed to the fixed size thread block
|
13 years ago |
|
Marina Kolpakova
|
3c2691ef96
|
LBP: minor changes
|
13 years ago |
|
Marina Kolpakova
|
ed1b293d34
|
refactored GPU LBP cascade. Added support for big images. Fixed bug in connected components function
|
13 years ago |
|
Marina Kolpakova
|
86d785622b
|
LBP: switched to texture implementation
|
13 years ago |
|
Marina Kolpakova
|
b0606b0557
|
LBP classifer moved to ptr from DevMem2D
|
13 years ago |
|
Marina Kolpakova
|
436d2ff1fc
|
LBP classifer moved to ptr from DevMem2D
|
13 years ago |
|
Marina Kolpakova
|
f094983d49
|
reduced the number of registers used in the LBP responce operator
|
13 years ago |
|
Marina Kolpakova
|
09269b4cd8
|
fixed backward compatibility with less than 1.2 CUDA capability
|
13 years ago |
|
Andrey Kamaev
|
e900952b87
|
quiet most of the OSX build warnings
|
13 years ago |
|
Marina Kolpakova
|
bc83011736
|
removed unnecessary memory allocation in LBP classifier
|
13 years ago |
|
Marina Kolpakova
|
eb599f6832
|
added checking destination matrix size for resize and integral functions
|
13 years ago |
|
Marina Kolpakova
|
6c0af66fd8
|
removed unused class filds
|
13 years ago |
|
Marina Kolpakova
|
e496345d33
|
added lbp cascade test, fixed race conditions problems
|
13 years ago |
|
Marina Kolpakova
|
248f39e13e
|
added performance test for LBP classifier.
|
13 years ago |
|
Marina Kolpakova
|
4128d5782f
|
added hipotesis filtration
|
13 years ago |
|
Marina Kolpakova
|
a53f0f397e
|
LBP classifier: tracking of count of detected objects was moved in cascadeclassifier.cpp
|
13 years ago |
|
Marina Kolpakova
|
cdbda45125
|
merged GPU resize became same as CPU implementation
|
13 years ago |
|
Marina Kolpakova
|
76354287aa
|
fixed build without CUDA
|
13 years ago |
|
Marina Kolpakova
|
319c20c797
|
result storing: atomic based
|
13 years ago |
|
Marina Kolpakova
|
a9f2f522e7
|
LBP classifier was refactored, added parameter for max size of detected object
|
13 years ago |
|
Marina Kolpakova
|
e6f7e4d83e
|
GPU resize became same as CPU implementation
|
13 years ago |
|
Marina Kolpakova
|
60b73e7471
|
gpu classifier default values became same as to cpu implementation
|
13 years ago |
|
Marina Kolpakova
|
e7f5978768
|
classifier loading has been improved
|
13 years ago |
|
Andrey Kamaev
|
afe11f69fb
|
Fixed Windows build warnings and configuration problem.
|
13 years ago |
|
Marina Kolpakova
|
304dac7f00
|
fixed #2113
|
13 years ago |
