[GraduationProject]压箱底的Qt如何捞出来之后晾干了再用！

去年暑假用QT做了一件简单的基于SIFT+RANSAC的拼接小程序，当时用的opencv版本是2.4.10.

一年过去。。。。做毕设的时候再次用到，电脑上opencv的版本已经到了3.3.0.

从新在QT上配置opencv经历了一整天的持续作战。。。。

首先参考如下教程用mingw编译适合qt的动态库：

https://blog.csdn.net/qq_17550379/article/details/78277313

opencv3.3的cmake过程真的非常多雷区啊！！最后生成的时候因为opencv_hdf和电脑上的Anaconda好像有冲突，所以这个库就没有编译。

编译好之后在qt中的.pro文件里填好

INCLUDEPATH += E:/OpenCV3.3_Qt/install/include  \
               E:/OpenCV3.3_Qt/install/include/opencv  \
               E:/OpenCV3.3_Qt/install/include/opencv2

LIBS += -LE:/OpenCV3.3_Qt/install/x86/mingw/lib  \
-llibopencv_aruco330  \
-llibopencv_bgsegm330  \
-llibopencv_bioinspired330  \
-llibopencv_calib3d330  \
-llibopencv_ccalib330  \
-llibopencv_core330  \
-llibopencv_datasets330  \
-llibopencv_dnn330  \
-llibopencv_dpm330  \
-llibopencv_face330  \
-llibopencv_features2d330  \
-llibopencv_flann330  \
-llibopencv_highgui330  \
-llibopencv_img_hash330  \
-llibopencv_imgcodecs330  \
-llibopencv_imgproc330  \
-llibopencv_line_descriptor330  \
-llibopencv_ml330  \
-llibopencv_objdetect330  \
-llibopencv_optflow330  \
-llibopencv_phase_unwrapping330  \
-llibopencv_photo330  \
-llibopencv_reg330  \
-llibopencv_rgbd330  \
-llibopencv_saliency330  \
-llibopencv_shape330  \
-llibopencv_stereo330  \
-llibopencv_stitching330  \
-llibopencv_structured_light330  \
-llibopencv_superres330  \
-llibopencv_surface_matching330  \
-llibopencv_text330  \
-llibopencv_tracking330  \
-llibopencv_video330  \
-llibopencv_videoio330  \
-llibopencv_videostab330  \
-llibopencv_xfeatures2d330  \
-llibopencv_ximgproc330  \
-llibopencv_xobjdetect330  \
-llibopencv_xphoto330

这时看起来编译是都能通过的，但是程序总是异常退出 T_T

陆陆续续查了很多资料，这篇博文给了我启发：

https://blog.csdn.net/qq_21078557/article/details/78962135

QT中是可以自己设定build时的环境的！！！

于是就发现

这个PATH这里果然保存的最后一条还是opencv2.4.10时的路径！！

改成今天mingw编译之后和lib文件夹一起的bin的路径就好辣！！

希望之后做界面成功！！！！