用github中的开源代码时,发现报了这样一段话:
OpenCV is built with OpenMP support. This usually results in poor performance. For details, see https://github.com/tensorpack/benchmarks/blob/master/ImageNet/benchmark-opencv-resize.py
以前从没听说过,于是进去看了看,发现是一个python下各安装版本opencv的测试。测试者在其中给出了自己的测试环境和测试结果。测试者根据自己的测试结果,得出了几个结论:
On E5-2680v3, archlinux, this script prints:
0.61s for system opencv 3.4.0-2.
>5 s for anaconda opencv 3.3.1 py36h6cbbc71_1.
On E5-2650v4, this script prints:
0.6s for opencv built locally with -DWITH_OPENMP=OFF
0.6s for opencv from pip install opencv-python
.
1.3s for opencv built locally with -DWITH_OPENMP=ON
2s for opencv from conda install
.
也就是说,测试者认为,使用pip安装的opencv要比conda装的快很多,编译时不用OpenMP的OpenCV要比用了的快很多。原因我也不懂,由于这个时间点我也没设备可以测试是否真是这样,所以暂且记在这里,如果以后确实有这个时间性能的需求再测试。
本文内容由网友自发贡献,版权归原作者所有,本站不承担相应法律责任。如您发现有涉嫌抄袭侵权的内容,请联系:hwhale#tublm.com(使用前将#替换为@)