[TOC]
背景
为了在某个conda环境使用带有cuda支持的opencv，手动编译
环境
ubuntu+cuda+conda+cmake
激活conda环境
conda create --name track python=3.6.9
conda activate track
建议安装上numpy，编译的opencv-python会有依赖
conda install numpy
下载
下载opencv和opencv_contrib包，并解压
wget https://github.com/opencv/opencv/archive/refs/tags/4.6.0.zip
wget https://github.com/opencv/opencv_contrib/archive/refs/tags/4.6.0.zip
mv 4.6.0.zip opencv_4.6.0.zip
mv 4.6.0.zip.1 opencv_contrib_4.6.0.zip
unzip opencv_4.6.0.zip
unzip opencv_contrib_4.6.0.zip
建立目录
cd opencv-4.6.0
mkdir build
cd build
编译
cmake
cmake -D CMAKE_BUILD_TYPE=RELEASE \
-D OPENCV_EXTRA_MODULES_PATH=../../opencv_contrib-4.6.0/modules \
-D OPENCV_ENABLE_NONFREE=ON \
-D WITH_OPENMP=ON \
-D WITH_TBB=ON \
-D OPENCV_EXTRA_EXE_LINKER_FLAGS=-latomic \
-D ENABLE_PRECOMPILED_HEADERS=OFF \
-D WITH_CUDA=ON \
-D ENABLE_FAST_MATH=1 \
-D CUDA_FAST_MATH=1 \
-D WITH_CUBLAS=1 \
-D CMAKE_INSTALL_PREFIX=$(python -c "import sys; print(sys.prefix)") \
-D PYTHON3_EXECUTABLE=$(which python) \
-D PYTHON3_INCLUDE_DIR=$(python -c "from distutils.sysconfig import get_python_inc; print(get_python_inc())") \
-D PYTHON3_PACKAGES_PATH=$(python -c "from distutils.sysconfig import get_python_lib; print(get_python_lib())") \
..
编译安装
make
make install