环境： Mac os linux

编译器：vscode

管理包：brew

安装包：

opencv
eigen
g2o
sophus

主要步骤

opencv

brew search opencv #可以搜索brew可以安装的opencv的version

brew install opencv@2 #选择要安装的opencv

brew info opencv #检查安装结果，若某个依赖包的后面没有✅，可以直接brew install安装这个包

安装完成之后会出现以下的提示，记得需要分别执行，这样可以把opencv的头文件和lib放到相应的位置以便随后配置

If you need to have opencv@2 first in your PATH run:
echo ‘export PATH=”/usr/local/opt/opencv@2/bin:$PATH”‘ >> ~/.zshrc

For compilers to find opencv@2 you may need to set:
export LDFLAGS=”-L/usr/local/opt/opencv@2/lib”
export CPPFLAGS=”-I/usr/local/opt/opencv@2/include”

For pkg-config to find opencv@2 you may need to set:
export PKG_CONFIG_PATH=”/usr/local/opt/opencv@2/lib/pkgconfig”
opencv@2 和 opencv@3有一些包有区别，所以有时候编译找不到其中一个包可以看一下版本是否有误

这样安装好的opencv的include的内容是在 /usr/local/opt/opencv@2/include

而link library是在当前文件夹 /usr/local/opt/opencv@2/lib

eigen
1
2
brew install eigen
这样安装好的opencv的include的内容是在 /usr/local/opt/eigen/include

而link library是在当前文件夹 /usr/local/opt/eigen/lib

Sophus

https://github.com/strasdat/Sophus

Sophus 是应用在二维和三维问题中的李群方法的一个包，可以直接git clone下载然后按照网页提示步骤安装。

git clone https://github.com/strasdat/Sophus.git #下载包到本地

cd Sophus #打开文件夹

mkdir build #创建build文件夹用于存放makefile

cmake .. #cmake上级的 cmakelist.txt

make #运行上一步在build文件夹中生成的makefile

sudo install make #在 /usr/local/include 中安装sophus的头文件之类的文件

这样安装好的sophus的include的内容是在 /usr/local/include/sophus

而link library是在当前文件夹 ../sophus (就一个lib文件)

在 cmake .. 这一步下，可能会出现以下错误：

1
2
3

/Users/wangcaimeng/Sophus/sophus/so2.cpp:32:24: error: expression is not assignable unit_complex_.real() = 1.;
/Users/wangcaimeng/Sophus/sophus/so2.cpp:33:24: error: expression is not    assignable unit_complex_.imag() = 0.; 
~~~~~~~~~~~~~~~~~~~~ ^ 2 errors generated. make[2]: *** [CMakeFiles/Sophus.dir/sophus/so2.cpp.o] Error 1 make[1]: *** [CMakeFiles/Sophus.dir/all] Error 2 make: *** [all] Error 2

解决方案是：

将 so2.cpp 文件的32和33行改为：

unit_complex_.real(1.);
unit_complex_.imag(0.);
参考自：https://github.com/gaoxiang12/slambook/issues/56#issuecomment-334946067

g2o

https://github.com/RainerKuemmerle/g2o

g2o是brewsci中的一个用于非线性优化的包，在官方git的readme.md文件中，提及到macos系统可以直接用
1
brew install brewsci/science/g2o
来安装相关的包，但是我在安装的的过程中会涉及到 Cannot tap brewsci/science 的报错信息，这个问题没有查到明确的解法，所以就按照 sophus相同的办法安装了g2o。在官方的g2o文件的Compilation 部分也有具体提到安装的步骤：
1
2
3
4
5
6
7
8
9
10
11
git clone https://github.com/RainerKuemmerle/g2o.git

cd g2o

mkdir build

cmake ..

make

sudo install make #在 /usr/local/include 中安装g2o的头文件之类的文件
这样安装好的g2o的include的内容是在 /usr/local/include/g2o

而link library是在当前文件夹 ../g2o/lib

由此可得，基于brew install 安装的包都比较统一地把include文件和lib文件放在 /usr/local/opt/opencv@2/include 和 /usr/local/opt/opencv@2/lib中，但是直接从git下载并且需要cmake的包需要自行确定include和lib文件夹在哪里。

如果在安装过后都使用了 sudo install make 那么include的包会被安装在 /usr/local/include/ 下

vscode 配置

vscode的配置文件又四个，分别是: launch.json, tasks.json, settngs.json, c_cpp_properties.json

lauch.json

{
    // Use IntelliSense to learn about possible attributes.
    // Hover to view descriptions of existing attributes.
    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
    "version": "0.2.0",
    "configurations": [
        {
            "name": "(lldb) Launch",
            "type": "cppdbg",
            "request": "launch",
            "program": "${fileDirname}/a.out",
            "args": [],
            "stopAtEntry": true,
            "cwd": "${workspaceFolder}",
            "environment": [],
            "externalConsole": false,
            "MIMode": "lldb"
        }
    ]
}

settings.json

//settings.json
{
    "python.pythonPath": "usr/bin/python",
    "code-runner.executorMap": {
        "c": "cd $dir && make  && ./$fileNameWithoutExt && make clean",
        "cpp": "cd $dir && make  && ./$fileNameWithoutExt && make clean"
    },
    "editor.renderWhitespace": "all",
    "editor.renderLineHighlight": "all",
    "editor.formatOnSave": true,
    "code-runner.runInTerminal": true,
    "code-runner.ignoreSelection": true,
    "code-runner.enableAppInsights": false,
    "C_Cpp.updateChannel": "Insiders",
    "[makefile]": {
        "editor.insertSpaces": true
    },
    "C_Cpp.default.includePath": [
        "/usr/local/Cellar/opencv@2/2.4.13.7_11/include"
    ]
}

tasks.json

{
    "tasks": [
        {
            "type": "shell",
            "label": "Build",
            "command": "g++",
            "args": [
                //"main.cpp",
                "${file}",
                "-o",
                "${fileDirname}/${fileBasenameNoExtension}.out",
                "-I",
                "/usr/local/Cellar/opencv@2/2.4.13.7_11/include/opencv",
                "-I",
                "/usr/local/Cellar/opencv@2/2.4.13.7_11/include",
                "-L",
                "/usr/local/Cellar/opencv@2/2.4.13.7_11/lib",
                "-l",
                "opencv_stitching",
                "-l",
                "opencv_superres",
                "-l",
                "opencv_videostab",
                "-l",
                "opencv_aruco",
                "-l",
                "opencv_bgsegm",
                "-l",
                "opencv_bioinspired",
                "-l",
                "opencv_ccalib",
                "-l",
                "opencv_dnn_objdetect",
                "-l",
                "opencv_dpm",
                "-l",
                "opencv_face",
                "-l",
                "opencv_fuzzy",
                "-l",
                "opencv_hfs",
                "-l",
                "opencv_img_hash",
                "-l",
                "opencv_line_descriptor",
                "-l",
                "opencv_optflow",
                "-l",
                "opencv_reg",
                "-l",
                "opencv_rgbd",
                "-l",
                "opencv_saliency",
                "-l",
                "opencv_stereo",
                "-l",
                "opencv_structured_light",
                "-l",
                "opencv_phase_unwrapping",
                "-l",
                "opencv_surface_matching",
                "-l",
                "opencv_tracking",
                "-l",
                "opencv_datasets",
                "-l",
                "opencv_dnn",
                "-l",
                "opencv_plot",
                "-l",
                "opencv_xfeatures2d",
                "-l",
                "opencv_shape",
                "-l",
                "opencv_video",
                "-l",
                "opencv_ml",
                "-l",
                "opencv_ximgproc",
                "-l",
                "opencv_xobjdetect",
                "-l",
                "opencv_objdetect",
                "-l",
                "opencv_calib3d",
                "-l",
                "opencv_features2d",
                "-l",
                "opencv_highgui",
                "-l",
                "opencv_videoio",
                "-l",
                "opencv_imgcodecs",
                "-l",
                "opencv_flann",
                "-l",
                "opencv_xphoto",
                "-l",
                "opencv_photo",
                "-l",
                "opencv_imgproc",
                "-l",
                "opencv_core",
                "-g"

            ],
            
            "group": {
                "kind": "build",
                "isDefault": true
            },
            "problemMatcher": [
                "$gcc"
            ]
        }
    ],
    "version": "2.0.0"
}

c_cpp_properties.json

{
    "configurations": [
        {
            "name": "Mac",
            "includePath": [
                "${workspaceFolder}/**",
                "/usr/local/opt/opencv@2/include",
                "/usr/local/Cellar/opencv@2/2.4.13.7_11/include"
            ],
            "defines": [],
            "macFrameworkPath": [],
            "compilerPath": "/usr/bin/g++",
            "cStandard": "c11",
            "cppStandard": "c++11",
            "intelliSenseMode": "clang-x64",
            "browse": {
                "path": [
                    "/usr/local/Cellar/opencv@2/2.4.13.7_11/include"
                ],
                "limitSymbolsToIncludedHeaders": true,
                "databaseFilename": ""
            }
        }
    ],
    "version": 4
}

cmakelist 配置

cmake中主要配置的是 INCLUDE_DIRECTORIES , link_directoryies

注意着两部分的路径药分别和之前安装的不同的包的路径相匹配，这样就可以运行cmake进行编译了

光

Mac使用brew在vscode下安装并配置【opencv】【g2o】【sophus】

vscode 配置

cmakelist 配置