Building image processing platform based on QT and OPENCV under Windows platform

    in previous blogs, the image processing framework based on VS and OpenCV under the Windows platform and the QT and OPENCV based on Linux platform were described in more detail, and the corresponding free video was generated. The main content of this blog is based on the latest version of the corresponding tools, under the Windows platform, "based on QT and OPENCV building image processing platform", and further research on how to make the image processing based on QT WYSIWYG convenience, and finally a comparison with VS, to make a preliminary summary. mainly divided into 3 parts, one is the current mode, WINDOWS+QT environment, two is a companion version of the OPENCV recompile, three is a combination of the environment to configure the success. first, build the QT environment under the Windows platformlatest version of QT for (December 5, 2017 version)

For (may need to register an account):

Http://download.qt.io/official_releases/qt/5.10/5.10.0/qt-opensource-windows-x86-5.10.0.exeThe most prominent feeling is that in the latest version of QT, a high degree of integration, the pre-build version of the various environments are published together. If you already have vs2013/vs2015 installed on your machine, you can choose the corresponding version. In this article, Imultipleafter the attempt, feel the top select "MingW" This version, while casting tools in the "MingW", so that the one-time installation in place. (select only these two items)in this case, to install QT, configure kit, you can basically run the form directly. Note that both the compiler and the debugger are filled out correctly. Second, supporting OPENCV compilation    in order to work with the current QT program, the OPENCV must be recompiled. The tools used are still cmakeSelect MinGW makefiles, select "Specify Native compilers"fill in the C and C + + compiler addresses separately, this is related to the address you installed QT, my address can be used as a reference:C Correspondence: D:/qt/qt5.10.0/tools/mingw530_32/bin/gcc.exeC + + correspondence:D:/qt/qt5.10.0/tools/mingw530_32/bin/g++.exeat this point, the finish will be error, pay attention to open "advanced", and then set Cmake_make_program to your Mingw32-make address, my address for referenceD:/qt/qt5.10.0/tools/mingw530_32/bin/mingw32-make.exe Click Configure to start the configuration. Note that the "Build_opencv_world" mode cannot be selected and can only be compiled separately. (For specific reasons not detailed, if the person know please drink me contact ", if you want to add contribute, please also set the correct.) Also, be sure to choose WITH_QTClick Configure, will report some errors, mainly is the QT configuration problem. where the address of the QMake.exe is correctly configured, and follow the similar "D:\Qt\Qt5.10.0\5.10.0\mingw53_32\include "The pairing address is set correctlyfinally configure no problem, click Generate. In this process, if the report lacks "Libwinpthread-1.dll ", you need to download one from the web and place it in the directory where path is located (such as a 64-bit system in SysWow64, 32-bit system in System32), which can be generated correctly. Enter the code generation directory, run "Mingw32-make" and "Mingw32-make install" to generate results, it will take about 1 hours. If you do not have path set, you can refer to this method in my. Two 100%, which represents the success of the build.    third, the production of the basic framework.     this time, it should be relatively simple. Mainly the configuration file and header file problems, simply paste it:add a header address to the configuration file:environment variable inside the bin added to theAdd related code:ok! Thanks for reading this, hope to help!

Building image processing platform based on QT and OPENCV under Windows platform