【opencv3.3】VS2015+opencv3.3 GPU模块编译（包含opencv_contrib模块）

据官方说法，目前还不是太稳定的算法模块都在opencv_contrib里边，由于不稳定，所以不能在release版本里发行，只有在稳定以后才会放进release里边。但是这里边有很多我们经常要用的算法，比如SIFT，SURF等（在xfeatures2d 模块里边）。官网提供了说明，可以把opencv_contrib扩展模块添加编译到已安装的opencv3里边。

同时我们还需要编译opencv的GPU模块，以便在GPU上加速执行这些算法。

1.点[Brouse Source…]，选择OpenCV源码那个sources文件夹的路径。 
点[Brouse Build…]，选择要生成的工程的路径。如下图：

2.点击 [Configure]，出现对话框说文件夹不存在要不要新建文件夹，点yes，然后出现对话框选择生成的工程版本，如下图：

3.点[Finished]，一段读条后会生成工程。完成后可以检查一下窗口下部的框，如果正确安装、配置CUDA，应该会有如下字样 CUDA detected+版本号：

4.检查一下WITH_CUDA选项，如果Cmake检测到你安装了CUDA，应该是自动勾上的。如果没自动勾上那就把它勾上。如下图：

5.下载opencv_contrib模块，链接：https://github.com/opencv/opencv_contrib

解压后，我把它放到了opencv3.3的目录下

6.在cmake界面找到OPENCV_EXTRA_MODULES_PATH，修改其值为：D:/opencv3.3/opencv_contrib-master/modules，就是第5步中modules的路径

7.确认好选项之后再按[Configure]

8.按[Generate]生成工程，如果配置和生成工程完全ok底下应该有Configuring done和Generating done两行。如下图： 

9.关掉cmake，关掉杀毒软件，在build_opencv3.3_contrib找到OpenCV.sln用VS2015打开，选择生成-重新生成解决方案。

等待两个多小时……

10.编译好后，找到解决方案目录里的[CMakeTargets]项展开的[INSTALL]项，右键->[Project Only(仅项目)]->[Build Only INSTALL(仅生成INSTALL)]。这时在D:\build_opencv3.3_contrib\install\x64\vc14生成了编译好的库（默认生成debug的库，修改为release编译生成release的库）。

添加path环境变量：D:\build_opencv3.3_contrib\install\x64\vc14\bin

11.测试gpu模块与opencv_contrib模块。

新建VS2015控制台应用程序，

工程属性--配置属性--VC++目录--包含目录  中添加：

D:\build_opencv3.3_contrib\install\include\opencv;

D:\build_opencv3.3_contrib\install\include\opencv2;

D:\build_opencv3.3_contrib\install\include;

工程属性--配置属性--VC++目录--库目录  中添加：

D:\build_opencv3.3_contrib\install\x64\vc14\lib

工程属性--配置属性--链接器--输入--附加依赖项 中添加：

opencv_aruco331d.lib
opencv_bgsegm331d.lib
opencv_bioinspired331d.lib
opencv_calib3d331d.lib
opencv_ccalib331d.lib
opencv_core331d.lib
opencv_cudaarithm331d.lib
opencv_cudabgsegm331d.lib
opencv_cudacodec331d.lib
opencv_cudafeatures2d331d.lib
opencv_cudafilters331d.lib
opencv_cudaimgproc331d.lib
opencv_cudalegacy331d.lib
opencv_cudaobjdetect331d.lib
opencv_cudaoptflow331d.lib
opencv_cudastereo331d.lib
opencv_cudawarping331d.lib
opencv_cudev331d.lib
opencv_datasets331d.lib
opencv_dnn331d.lib
opencv_dpm331d.lib
opencv_face331d.lib
opencv_features2d331d.lib
opencv_flann331d.lib
opencv_fuzzy331d.lib
opencv_highgui331d.lib
opencv_img_hash331d.lib
opencv_imgcodecs331d.lib
opencv_imgproc331d.lib
opencv_line_descriptor331d.lib
opencv_ml331d.lib
opencv_objdetect331d.lib
opencv_optflow331d.lib
opencv_phase_unwrapping331d.lib
opencv_photo331d.lib
opencv_plot331d.lib
opencv_reg331d.lib
opencv_rgbd331d.lib
opencv_saliency331d.lib
opencv_shape331d.lib
opencv_stereo331d.lib
opencv_stitching331d.lib
opencv_structured_light331d.lib
opencv_superres331d.lib
opencv_surface_matching331d.lib
opencv_text331d.lib
opencv_tracking331d.lib
opencv_video331d.lib
opencv_videoio331d.lib
opencv_videostab331d.lib
opencv_xfeatures2d331d.lib
opencv_ximgproc331d.lib
opencv_xobjdetect331d.lib
opencv_xphoto331d.lib

可以在D:\opencv3.3\sources\samples\gpu\surf_keypoint_matcher.cpp找到一个测试例程，对其代码做简单修改如下：

#include <iostream>#include "opencv2/opencv_modules.hpp"#include "opencv2/core.hpp"#include "opencv2/features2d.hpp"#include "opencv2/highgui.hpp"#include "opencv2/cudafeatures2d.hpp"#include "opencv2/xfeatures2d/cuda.hpp"using namespace std;using namespace cv;using namespace cv::cuda;int main(){GpuMat img1, img2;img1.upload(imread("1.bmp", IMREAD_GRAYSCALE));img2.upload(imread("2.bmp", IMREAD_GRAYSCALE));cv::cuda::printShortCudaDeviceInfo(cv::cuda::getDevice());SURF_CUDA surf;// detecting keypoints & computing descriptorsGpuMat keypoints1GPU, keypoints2GPU;GpuMat descriptors1GPU, descriptors2GPU;surf(img1, GpuMat(), keypoints1GPU, descriptors1GPU);surf(img2, GpuMat(), keypoints2GPU, descriptors2GPU);cout << "FOUND " << keypoints1GPU.cols << " keypoints on first image" << endl;cout << "FOUND " << keypoints2GPU.cols << " keypoints on second image" << endl;// matching descriptorsPtr<cv::cuda::DescriptorMatcher> matcher = cv::cuda::DescriptorMatcher::createBFMatcher(surf.defaultNorm());vector<DMatch> matches;matcher->match(descriptors1GPU, descriptors2GPU, matches);// downloading resultsvector<KeyPoint> keypoints1, keypoints2;vector<float> descriptors1, descriptors2;surf.downloadKeypoints(keypoints1GPU, keypoints1);surf.downloadKeypoints(keypoints2GPU, keypoints2);surf.downloadDescriptors(descriptors1GPU, descriptors1);surf.downloadDescriptors(descriptors2GPU, descriptors2);// drawing the resultsMat img_matches;drawMatches(Mat(img1), keypoints1, Mat(img2), keypoints2, matches, img_matches);namedWindow("matches", 0);imshow("matches", img_matches);waitKey(0);return 0;}

测试图片：

运行结果：