ubuntu22.04编译安装opencv和opencv_contrib_ubuntu22.04安装opencv

1. 下载OpenCV和OpenCV源文件

opencv: https://opencv.org/releases/
opencv_contrib: https://github.com/opencv/opencv_contrib
注意opencv_contrib和opencv版本要对应
下载后将安装包解压到主目录，重命名为opencv和opencv_contrib

2. 安装依赖项

安装g++, cmake, make, wget, unzip

sudo apt install -y g++ 
sudo apt install -y cmake 
sudo apt install -y make 
sudo apt install -y wget 
sudo apt install -y unzip
1
2
3
4
5

安装opencv依赖的基本库

sudo apt-get install build-essential libgtk2.0-dev libgtk-3-dev libavcodec-dev libavformat-dev libjpeg-dev libswscale-dev libtiff5-dev
1

3. 编译安装

如果提前安装了anaconda， 建议先把anaconda环境变量注释掉，并把anaconda文件夹移进回收站等装完再移出来（改软连接也行，但上述更无脑些）

主目录下打开终端

1. 新建build

cd opencv
mkdir build
cd build
1
2
3

2. cmake配置

 cmake -DCMAKE_BUILD_TYPE=RELEASE \
    -DCMAKE_INSTALL_PREFIX=/usr/local \
    -DOPENCV_EXTRA_MODULES_PATH=~/opencv_contrib/modules .. \
    -DWITH_CUDA=1 \
    -DCUDA_ARCH_BIN=8.6 \
    -DENABLE_FAST_MATH=1 \
    -DCUDA_FAST_MATH=1 \
    -DWITH_CUBLAS=1 \
    -DOPENCV_GENERATE_PKGCONFIG=1 \
    ..
1
2
3
4
5
6
7
8
9
10

注意：上面的DCUDA_ARCH_BIN指定GPU算力，在NVIDIA官网查询 https://developer.nvidia.com/cuda-gpus#compute，常用的30系显卡是8.6，40系显卡是8.9

3. make编译
   可以先查一下最多用多少线程

nproc
1

比如输出 16
那么就

make -j16
1

4. 用make进行安装

sudo make install 
1

4. 环境配置

1. 配置pkg-config环境

sudo gedit /etc/profile.d/pkgconfig.sh
1

可能为空文件，在文件中加入

export PKG_CONFIG_PATH=/usr/local/lib/pkgconfig:$PKG_CONFIG_PATH
1

保存并退出，source激活

source /etc/profile
1

验证是否成功

pkg-config --libs opencv4
1

输出类似下面这样即为成功

-L/usr/local/lib -lopencv_gapi -lopencv_stitching -lopencv_alphamat -lopencv_aruco -lopencv_barcode -lopencv_bgsegm -lopencv_bioinspired -lopencv_ccalib -lopencv_cudabgsegm -lopencv_cudafeatures2d -lopencv_cudaobjdetect -lopencv_cudastereo -lopencv_dnn_objdetect -lopencv_dnn_superres -lopencv_dpm -lopencv_face -lopencv_freetype -lopencv_fuzzy -lopencv_hfs -lopencv_img_hash -lopencv_intensity_transform -lopencv_line_descriptor -lopencv_mcc -lopencv_quality -lopencv_rapid -lopencv_reg -lopencv_rgbd -lopencv_saliency -lopencv_stereo -lopencv_structured_light -lopencv_phase_unwrapping -lopencv_superres -lopencv_cudacodec -lopencv_surface_matching -lopencv_tracking -lopencv_highgui -lopencv_datasets -lopencv_text -lopencv_plot -lopencv_videostab -lopencv_cudaoptflow -lopencv_optflow -lopencv_cudalegacy -lopencv_videoio -lopencv_cudawarping -lopencv_wechat_qrcode -lopencv_xfeatures2d -lopencv_shape -lopencv_ml -lopencv_ximgproc -lopencv_video -lopencv_xobjdetect -lopencv_objdetect -lopencv_calib3d -lopencv_imgcodecs -lopencv_features2d -lopencv_dnn -lopencv_flann -lopencv_xphoto -lopencv_photo -lopencv_cudaimgproc -lopencv_cudafilters -lopencv_imgproc -lopencv_cudaarithm -lopencv_core -lopencv_cudev
1

2. 配置动态库环境

sudo gedit /etc/ld.so.conf.d/opencv4.conf
1

文件末尾加上

/usr/local/lib
1

使配置生效

sudo ldconfig
1

5. demo测试

打开新终端，进入到example_cmake目录

cd ~/opencv/samples/cpp/example_cmake
1

cmake .
make
./opencv_example
1
2
3

正常来说，应该有摄像头窗口界面；如果没有摄像头，出现“Hello OpenCV“也测试成功