opencv3.2 安装说明书
软件环境
ubuntu 14,04 python2.7 cmake , g++, gcc
1 安装OpenCV依赖
1.1 升级系统
$ sudo apt-get update
$ sudo apt-get upgrade
1.2 安装编译工具
由于是通过源代码安装 , 说以需要安装编译工具.
$ sudo apt-get install build-essential cmake pkg-config
1.3 安装依赖库
由于OpenCV是计算机视觉库，总需要加载一些图像文件（ JPEG, PNG, TIFF）。使用下面命令安装一些必要的图像库：
$ sudo apt-get install libjpeg8-dev libtiff5-dev libjasper-dev libpng12-dev
除了图片之外，OpenCV还要处理视频文件。使用下面命令安装一些视频编解码库
sudo apt-get install libavcodec-dev libavformat-dev libswscale-dev libv4l-dev
sudo apt-get install libxvidcore-dev libx264-dev
OpenCV的GUI模块highgui依赖Gtk。安装gtk库：
sudo apt-get install libgtk-3-dev
下面安装一些可以提高OpenCV性能的库，如矩阵操作:
sudo apt-get install libatlas-base-dev gfortran
安装Python2和Python3的开发库：
sudo apt-get install python2.7-dev python3.5-dev
PS Ubuntu14.04中含有python2.7 , 如果不想自己的系统环境太复杂,可以选择不去安装python2.7
下载opencv 源代码
目前OpenCV的最新稳定版是3.2 , 可以去opencv官网上下载, 也可以去github总下载最新版本
最好再下载opencv_contrib,
opencvcontrib模块是Opencv 没有经过足够测试的的新功能模块.OpenCV 在发布新的版本之前,首先在opencvcontrib仓库中发布。当模块成熟并获得普及后，便转移到中央OpenCV存储库，并发布新的版本; :)
注意：opencv和opencv_contrib的版本要一致
# 由于我下载的opencv3.2 所以会在github下找到相应的版本
wget https://github.com/opencv/opencv_contrib/archive/3.2.0.tar.gz
tar -zvxf 3.2.0.tar.gz # 解压以后生成 opencv_contrib-3.2.0
2. 编译OpenCV
切换到opencv 的build 目录 , opencvbuilddirectory
cd ## 跳转到opencv 目录
mkdir build ## 创建 opencv_build_directory
cd ## 切换到 opencv_build_directory 目录
生成编译文件
cmake -D CMAKE_BUILD_TYPE=RELEASE \
-D CMAKE_INSTALL_PREFIX=/usr/local \
-D INSTALL_PYTHON_EXAMPLES=ON \
-D INSTALL_C_EXAMPLES=OFF \
-D OPENCV_EXTRA_MODULES_PATH=/home/linux/work_space/opencv_contribute/opencv_contrib-3.2.0/modules \
-D PYTHON_EXECUTABLE=/usr/bin/python2.7 -D BUILD_EXAMPLES=ON ..
make -j #NUM 自己设定 我设置的是80
ps: 我这里opencvcontrib 的目录是/home/linux/workspace/opencvcontribute/opencvcontrib-3.2.0/ ; 所以参数写成 -D OPENCVEXTRAMODULESPATH=/home/linux/workspace/opencvcontribute/opencvcontrib-3.2.0/modules
最后的 .. 是由于build 目录在openCV里面, 我想编译build 外面的文件,所以要用相对路径 ..
如果是MacOS 可以采用以下方式编译:
cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local \
-D PYTHON2_PACKAGES_PATH=/usr/local/lib/python2.7/site-packages \
-D PYTHON2_LIBRARY=/Library/Frameworks/Python.framework/Versions/2.7/bin \
-D PYTHON2_INCLUDE_DIR=/Library/Frameworks/Python.framework/Headers \
-D INSTALL_C_EXAMPLES=ON -D INSTALL_PYTHON_EXAMPLES=ON \
-D BUILD_EXAMPLES=ON \
-D OPENCV_EXTRA_MODULES_PATH=/Users/lee/workspace/opencv_contrib-3.2.0/modules ..
参数说明:
There are somevery importantoptions we are supplying to CMake here, so let’s break them down:
CMAKE_BUILD_TYPE: This option indicates that we are building a release binary of OpenCV.
CMAKE_INSTALL_PREFIX: The base directory where OpenCV will be installed.
PYTHON2_PACKAGES_PATH: Theexplicitpath to where oursite-packagesdirectory lives in ourcvvirtual environment.
PYTHON2_LIBRARY: Path to our Hombrew installation of Python.
PYTHON2_INCLUDE_DIR: The path to our Python header files for compilation.
INSTALL_C_EXAMPLES: Indicate that we want to install the C/C++ examples after compilation.
INSTALL_PYTHON_EXAMPLES: Indicate that we want to install the Python examples after complication.
BUILD_EXAMPLES: A flag that determines whether or not the included OpenCV examples will be compiled or not.
OPENCV_EXTRA_MODULES_PATH:This option is extremely important— here we supply the path to theopencv_contribrepo that we pulled down earlier, indicating that OpenCV should compile the extra modules as well.
3 安装
make install # 安装
3.1 Mac OS
PS: 如果你是MacOS 那么还需要手动的添加cv2的相关文件到python 中;
我这里提供两种方式:
如果使用brew 安装的, (安装步骤详见 )仅需如此:
查看有没有这个目录 /usr/local/Cellar/opencv/2.4.10/
cd /Library/Python/2.7/site-packages/
ln -s /usr/local/Cellar/opencv/2.4.13.2/lib/python2.7/site-packages/cv.py cv.py
ln -s /usr/local/Cellar/opencv/2.4.13.2/lib/python2.7/site-packages/cv2.so cv2.so
如果是安装以上的源码方式在本地编译安装的情况:
sudo ln -s /usr/local/Cellar/opencv/2.4.13.2/lib/python2.7/site-packages/cv2.so
/Library/Python/2.7/site-packages/cv2.so
即可
