I think this is due to the use of CUDA, but (gcc) does not support CUDA.
This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
I think this is due to the use of CUDA, but (gcc) does not support CUDA.
| | 2 | retagged |
I think this is due to the use of CUDA, but (gcc) does not support CUDA.
The same code. Windows 7; Intel Core i7-3612QM CPU 2.10 GHz (4 kernels 8 streams); RAM 8Gb; First example. VS 2017 OpenCV 3.4.0; Windows SDK version 10.0.16299.0 to target Windows 6.1.7601; The C compiler identification is MSVC 19.13.26131.1 Second example. The C compiler identification is GNU 7.3.0; Check for working C compiler: C: /msys64/mingw64/bin/gcc.exe; Found OpenCV: C: / msys64 / mingw64 (found version "3.4.0") In first example: time work for cv::matchTemplate(scene, img_templ, corr, CV_TM_CCORR_NORMED) ~= 0.6-0,7 s. In second example: time work for cv::matchTemplate(scene, img_templ, corr, CV_TM_CCORR_NORMED) ~= 1.1~1,4 s. I think this don't understand. Why is due to the use of CUDA, but (gcc) it different time? How does not support CUDA.it fix?
The same code. Windows 7; 7 x64; Intel Core i7-3612QM CPU 2.10 GHz (4 kernels 8 streams); RAM 8Gb; First example. VS 2017 OpenCV 3.4.0; Windows SDK version 10.0.16299.0 to target Windows 6.1.7601; The C compiler identification is MSVC 19.13.26131.1 Second example. The C compiler identification is GNU 7.3.0; Check for working C compiler: C: /msys64/mingw64/bin/gcc.exe; Found OpenCV: C: / msys64 / mingw64 (found version "3.4.0") In first example: time work for cv::matchTemplate(scene, img_templ, corr, CV_TM_CCORR_NORMED) ~= 0.6-0,7 s. In second example: time work for cv::matchTemplate(scene, img_templ, corr, CV_TM_CCORR_NORMED) ~= 1.1~1,4 s. I don't understand. Why is it different time? How does it fix?
