I use cmake-gui to build opencv on NVIDIA Jetsion Xavier NX, and tick WITH OPENMP option.After, I loop simple cv::resize (source image size:1280x1024, destination image size:10000x10000, INTER_LINRAR). However, I find that only one core of CPU is 100%. It seems that cv::parallel_for not use openmp in cv::resize(). The code seems to say that macro HAVE_OPENMP must exist. So how can I use parallel_for with openmp?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Multithreading issue when using parallel_for_ while running on ubuntu | 2 | 792 | March 30, 2022 | |
| Self-built paralle code is way slow then the sample code | 2 | 107 | May 9, 2024 | |
| OpenMP burning CPU cycles: cv::resize() is very slow unless cv::imshow() is used | 4 | 197 | June 15, 2024 | |
| Opencv parallel framework | 6 | 1103 | November 17, 2022 | |
| How to select to use OpenMP or Intel TBB? | 10 | 3158 | November 29, 2022 |