YOLO系列——Ubuntu20.04下通过conda虚拟环境源码安装opencv-4.10与opencv_contrib-4.10.0

YOLO系列------Ubuntu20.04下通过conda虚拟环境源码安装opencv-4.10与opencv_contrib-4.10.0

• 创建虚拟环境
• 安装常用包
• opencv源码下载
• 安装
• • 1.将opencv-4.10.0和opencv_contrib-4.10.0解压
  • 2.将opencv_contrib-4.10.0解压后的文件拷贝到opencv-4.10文件下
  • 3.进入opencv-4.10.0并创建build文件夹
  • [4. 编译](#4. 编译)
  • [5. 安装](#5. 安装)
  • 6.版本检查

创建虚拟环境

conda create -n opencv410env python=3.10

安装常用包

conda install numpy pandas matplotlib jupyter

opencv源码下载

1.官网下载

2.csdn下载

安装

1.将opencv-4.10.0和opencv_contrib-4.10.0解压

2.将opencv_contrib-4.10.0解压后的文件拷贝到opencv-4.10文件下

3.进入opencv-4.10.0并创建build文件夹

cd opencv-4.10.0
mkdir build
cd build

4. 编译

cmake -D CMAKE_BUILD_TYPE=RELEASE \
      -D CMAKE_INSTALL_PREFIX=/home/yy/miniconda3/envs/opencv410env \
      -D OPENCV_EXTRA_MODULES_PATH=/home/yy/software/opencv-4.10.0/opencv_contrib-4.10.0/modules \
      -D PYTHON3_EXECUTABLE=/home/yy/miniconda3/envs/opencv410env/bin/python \
      -D PYTHON3_INCLUDE_DIR=/home/yy/miniconda3/envs/opencv410env/include/python3.10 \
      -D PYTHON3_LIBRARY=/home/yy/miniconda3/envs/opencv410env/lib/libpython3.10.so \
      -D PYTHON3_NUMPY_INCLUDE_DIRS=/home/yy/miniconda3/envs/opencv410env/lib/python3.10/site-packages/numpy/_core/include \
      -D PYTHON3_PACKAGES_PATH=/home/yy/miniconda3/envs/opencv410env/lib/python3.10/site-packages \
      -D BUILD_opencv_python3=ON \
      -D BUILD_opencv_python2=OFF \
      -D BUILD_EXAMPLES=OFF \
      -D BUILD_TESTS=OFF \
      -D WITH_GTK=ON \
      -D WITH_FFMPEG=ON ..

路径需要做适当替换

5. 安装

make install

6.版本检查

opencv_version