方法有2：

1. 使用源代码CMake编译
2. 使用低版本opencv模块

本文主要记录方法2。3.4.2.16是最后一个支持nonfree模块的opencv库

1. 卸载旧版本opencv

   pip uninstall opencv-python
   pip uninstall opencv-contrib-python

2. opencv 3.4.2.16版本最多支持到python 3.6版本，需要创建虚拟环境

   conda install nb_conda
   conda create -n opencvnonfree python=3.6
   conda activate opencvnonfree

3. 安装所需的库 ```powershell pip install numpy pip install pandas pip install setuptools pip install —ignore-installed six pip install python-dateutil pip install jupyter

conda install imutils conda install scikit-image conda install —upgrade jupyterthemes

ref: https://blog.csdn.net/qq_26870933/article/details/88211711 https://blog.csdn.net/qq_30565883/article/details/79444750 https://www.zhihu.com/question/40012144

4. 安装opencv库
```powershell
pip install opencv-python==3.4.2.16
pip install opencv-contrib-python==3.4.2.16
ref:
https://blog.csdn.net/cliukai/article/details/102525486

1. 打开jupyter notebook

附上方法1的链接：https://opencv.org/releases/#https://sourceforge.net/projects/opencvlibrary/https://pypi.org/project/opencv-contrib-python/#fileshttps://xugaoxiang.com/2019/12/18/windows-10-cuda-cudnn/