How to build up OpenCV in Linux Back

  • Ubuntu 14.04 + OpenCV 3.1.0

1. Be sure to install cmake

sudo apt-get install cmake

2. Download and extract opencv

wget -O


3. Build

3.1 create a build directory

mkdir build && cd build

3.2 use cmake to build


  • Problem: qmake: could not exec '/usr/lib/i386-linux-gnu/qt4/bin/qmake': No such file or directory
  • Solution: install the qt dependence (sudo apt-get install qt4-qmake libqt4-dev)
  • Problem: cmake_cxx_compiler not set after enablelanguage
  • Solution: install c++ compiler (sudo apt-get install g++)
  • Problem:

    • The first "expected" hash should be always same (for specific OpenCV code version). The second "actual" hash is a signature of received data. And this received data is just broken due some reasons (firewall policy restrictions / other network problems / other software problems, like legacy CMake (2.8.7+ works fine) ): you may try to open file in "mc" editor and check it, sometimes there are messages like "403 forbidden", "timeout", "connection lost".
    • There are some discussion for the similar problem here (but about ffmpeg): #5546 #5895
  • Soluton:

    • disable this feature (-DWITH_IPP=OFF)
    • try to download these files manually and put on the right places. You can get these IPPICV files from here in the RAW mode. Script doesn't re-download file if it has right contents and it is located in the right place.
3.3 make

sudo make -j2

  • Note: this will take some time to complete.
  • Note: without ffmpeg, opencv cannot read a video
  • Problem: some lib*.so have undefined link to reference LibXML2
  • Solution: you should check the lib when installing ffmpeg. (eg. enable libbluray in ffmpeg will cause problem in ubuntu 14.04 when building opencv)
  • Problem: /usr/bin/ld: /usr/local/lib/libavcodec.a(avpacket.o): relocation R_X86_64_32 against `.rodata.str1.1' can not be used when making a shared object; recompile with -fPIC.
  • Solution: recompile libx264 with --enable-pic and --enable-shared because your system is 64-bit Ubuntu. (More details about installing ffmpeg on Ubuntu 12.04)
  • Problem: No rule to make target '/usr/lib/x86_64-linux-gnu/'
  • Solution: try to re-install libgl with sudo apt-get install --reinstall libgl1-mesa-dev
3.4 make install

sudo make install

4. Configure

sudo sh -c 'echo "/usr/local/lib" > /etc/'

sudo ldconfig

5. Reboot the computer

6. Create a project and run it

6.1 Create CMakeLists.txt

vim CMakeLists.txt

cmake_minimum_required(VERSION 2.8)
project( [projectName] )
find_package( OpenCV REQUIRED )
add_executable( [projectName] [projectName].cpp )
target_link_libraries( [projectName] ${OpenCV_LIBS} )
6.2 Run the project

cmake .



7. some problems about opencv in Visual Studio 2013

  • Problems: opencv_core249d.lib(array.obj) : error LNK2038: 檢測到 "RuntimeLibrary" 的不匹配项: 值 "MTd_StaticDebug" 不匹配值 "MDd_DynamicDebug"
  • Solution: 工程 -> 屬性 -> c/c++ -> 代碼生成 -> 運行庫 -> 多線程調試(/MTd)
  • Problems: 無法啟動此程序, 丟失 opencv_core249.dll
  • Solution: 將 opencv_core249.dll, opencv_highgui249.dll 複製到 exe 目錄下即可

8. build OpenCV with extra modles

  • build all modules
cd <opencv_build_directory>
cmake -DOPENCV_EXTRA_MODULES_PATH=<opencv_contrib>/modules <opencv_source_directory>
make -j5
  • build a specific module
 cmake -DOPENCV_EXTRA_MODULES_PATH=<opencv_contrib>/modules -DBUILD_opencv_legacy=OFF <opencv_source_directory>
Empty Comments
Sign in GitHub

As the plugin is integrated with a code management system like GitLab or GitHub, you may have to auth with your account before leaving comments around this article.

Notice: This plugin has used Cookie to store your token with an expiration.