opencv上gpu版surf特征点与orb特征点提取及匹配实例

标签：gpu版surfgpu orbsurf和orbgpu surf及orbsurf orb gpu -09-25 23:42464人阅读评论(0)收藏举报 分类：opencv（13）

版权声明：本文为博主原创文章，未经博主允许。

一、前言

本文主要实现了使用OpenCV里的gpu版surf特征检测器和gpu版orb检测器，分别对图片进行特征点提取及匹配，并对寻获的特征点进行了距离筛选，将匹配较为好的特征点进行展示

二、实现代码

我不生产代码，我只是代码的搬运工和修改工

[cpp]view plaincopy print? //main.cpp// #include<opencv2/core/core.hpp> #include<opencv2/imgproc/imgproc.hpp> #include<opencv2/highgui/highgui.hpp> #include<opencv2/gpu/gpu.hpp> #include<opencv2/nonfree/gpu.hpp> #include<opencv2/nonfree/features2d.hpp> #include<iostream> usingnamespacestd; usingnamespacecv; MatrotatedImage(constMat&_src,double_degree) { intwidth_src=_src.cols; intheight_src=_src.rows; floatcenter_x=width_src/2.0; floatcenter_y=height_src/2.0; doubleangle=_degree*CV_PI/180.; doublea=sin(angle),b=cos(angle); Matmap_matrix=getRotationMatrix2D(Point2f(center_x,center_y),_degree,1.0);//获得旋转矩阵 intheight_rotated=height_src*fabs(b)+width_src*fabs(a); intwidth_rotated=height_src*fabs(a)+width_src*fabs(b); map_matrix.at<double>(0,2)+=(width_rotated-width_src)/2.0;//将坐标移到中点 map_matrix.at<double>(1,2)+=(height_rotated-height_src)/2.0;//将坐标移到中点 Matdst; warpAffine(_src,dst,map_matrix,Size(width_rotated,height_rotated), CV_INTER_CUBIC|CV_WARP_FILL_OUTLIERS,BORDER_CONSTANT,cvScalarAll(0)); returndst; } //主要获得surf特征点、描述子、及特征点匹配 voidsurfExtractor(Mat&_src_Img,Mat&_dst_Img) { gpu::GpuMatsrc_gpu(_src_Img); gpu::GpuMatdst_gpu(_dst_Img); std::vector<KeyPoint>keypoints_src; std::vector<KeyPoint>keypoints_dst; std::vector<DMatch>matches; gpu::SURF_GPUFeatureFinder_gpu(500); gpu::GpuMatkeypoints_gpu_src,keypoints_gpu_dst; gpu::GpuMatdescriptors_gpu_src,descriptors_gpu_dst; std::vector<float>descriptors_v1,descriptors_v2; //计算特征点和特征描述子 FeatureFinder_gpu(src_gpu,gpu::GpuMat(),keypoints_gpu_src,descriptors_gpu_src); FeatureFinder_gpu(dst_gpu,gpu::GpuMat(),keypoints_gpu_dst,descriptors_gpu_dst); //将特征点下载回cpu，便于画图使用 FeatureFinder_gpu.downloadKeypoints(keypoints_gpu_src,keypoints_src); FeatureFinder_gpu.downloadKeypoints(keypoints_gpu_dst,keypoints_dst); //使用gpu提供的BruteForceMatcher进行特征点匹配 gpu::BruteForceMatcher_GPU<L2<float>>matcher_lk; matcher_lk.match(descriptors_gpu_src,descriptors_gpu_dst,matches,gpu::GpuMat()); floatmax_distance=0.2;//定义特征点好坏衡量距离 std::vector<DMatch>good_matches;//收集较好的匹配点 for(inti=0;i<descriptors_gpu_src.rows;i++){ if(matches[i].distance<max_distance){ good_matches.push_back(matches[i]); } } Matimage_matches; drawMatches(_src_Img,keypoints_src,_dst_Img,keypoints_dst,good_matches, image_matches,Scalar(0,255,0),Scalar::all(-1),vector<char>(),0); imshow("GpuSurf",image_matches); } voidorbExtractor(Mat&_src_Img,Mat&_dst_Img) { gpu::GpuMatsrc_gpu(_src_Img); gpu::GpuMatdst_gpu(_dst_Img); std::vector<KeyPoint>keypoints_src,keypoints_dst; gpu::GpuMatdescriptors_gpu_src,descriptors_gpu_dst; std::vector<DMatch>matches; gpu::ORB_GPUorb_finder(500); orb_finder.blurForDescriptor=true;//设置模糊 cv::gpu::GpuMatfullmask_1(src_gpu.size(),CV_8U,0xFF); cv::gpu::GpuMatfullmask_2(dst_gpu.size(),CV_8U,0xFF); orb_finder(src_gpu,fullmask_1,keypoints_src,descriptors_gpu_src); orb_finder(dst_gpu,fullmask_2,keypoints_dst,descriptors_gpu_dst); //使用gpu提供的BruteForceMatcher进行特征点匹配 gpu::BruteForceMatcher_GPU<HammingLUT>matcher_lk; matcher_lk.match(descriptors_gpu_src,descriptors_gpu_dst,matches,gpu::GpuMat()); floatmax_distance=60;//定义特征点好坏衡量距离 std::vector<DMatch>good_matches;//收集较好的匹配点 for(inti=0;i<descriptors_gpu_src.rows;i++){ if(matches[i].distance<max_distance){ good_matches.push_back(matches[i]); } } Matimage_matches; drawMatches(_src_Img,keypoints_src,_dst_Img,keypoints_dst,good_matches, image_matches,Scalar(255,0,0),Scalar::all(-1),vector<char>(),0); imshow("GpuORB",image_matches); } intmain() { intnum_devices=cv::gpu::getCudaEnabledDeviceCount(); if(num_devices<=0) { std::cerr<<"Thereisnodevice."<<std::endl; return-1; } intenable_device_id=-1; for(inti=0;i<num_devices;i++) { cv::gpu::DeviceInfodev_info(i); if(dev_info.isCompatible()) { enable_device_id=i; } } if(enable_device_id<0) { std::cerr<<"GPUmoduleisn'tbuiltforGPU"<<std::endl; return-1; } gpu::setDevice(enable_device_id); Matsrc_Img=imread("book.bmp",0); Matdst_Img=rotatedImage(src_Img,-30.0); surfExtractor(src_Img,dst_Img); orbExtractor(src_Img,dst_Img); cv::waitKey(0); return0; }

三、运行结果

运行环境为vs+opencv2.4.9+cuda7.0,结果展示如下，orb算法寻找特征点及计算描述子速度较快，gpu版的surf特征点对输入图片大小有要求，不能太小

使用GPU提取SURF局部特征

GPU做SURF特征提取的接口和CPU版本有所不同，不过这一部分可以完全参考<opencv_source_directory>/samples/gpu/surf_keypoint_matcher.cpp的例子代码。

我这里给出一个更加简化的例子，并添加一些中文注释和说明。

/*surf.cpp*/ #include <opencv2/imgcodecs.hpp>#include <opencv2/xfeatures2d/cuda.hpp>#include <opencv2/cudafeatures2d.hpp>using namespace std;int GetMatchPointCount(const char * pic_path_1,const char * pic_path_2) {/*指定使用的GPU序号，相关的还有下面几个函数可以使用cv::cuda::getCudaEnabledDeviceCount();cv::cuda::getDevice();cv::cuda::DeviceInfo*/cv::cuda::setDevice(0);/*向显存加载两张图片。这里需要注意两个问题：第一，我们不能像操作（主）内存一样直接一个字节一个字节的操作显存，也不能直接从外存把图片加载到显存，一般需要通过内存作为媒介第二，目前opencv的GPU SURF仅支持8位单通道图像，所以加上参数IMREAD_GRAYSCALE*/cv::cuda::GpuMat gmat1;cv::cuda::GpuMat gmat2;gmat1.upload(cv::imread(pic_path_1,cv::IMREAD_GRAYSCALE));gmat2.upload(cv::imread(pic_path_2,cv::IMREAD_GRAYSCALE));/*下面这个函数的原型是：explicit SURF_CUDA(double _hessianThreshold, //SURF海森特征点阈值int _nOctaves=4, //尺度金字塔个数int _nOctaveLayers=2, //每一个尺度金字塔层数bool _extended=false, //如果true那么得到的描述子是128维，否则是64维float _keypointsRatio=0.01f, bool _upright = false );要理解这几个参数涉及SURF的原理*/cv::cuda::SURF_CUDA surf(100,4,3); /*分配下面几个GpuMat存储keypoint和相应的descriptor*/cv::cuda::GpuMat keypt1,keypt2;cv::cuda::GpuMat desc1,desc2;/*检测特征点*/surf(gmat1,cv::cuda::GpuMat(),keypt1,desc1);surf(gmat2,cv::cuda::GpuMat(),keypt2,desc2);/*匹配，下面的匹配部分和CPU的match没有太多区别,这里新建一个Brute-Force Matcher，一对descriptor的L2距离小于0.1则认为匹配*/auto matcher=cv::cuda::DescriptorMatcher::createBFMatcher(cv::NORM_L2);vector<cv::DMatch> match_vec;matcher->match(desc1,desc2,match_vec);int count=0;for(auto & d:match_vec){if(d.distance<0.1) count++;}return count;}int main(int argc, const char* argv[]){GetMatchPointCount(argv[1],argv[2]);return 0;}123456789101112131415161718192223242526272829303132333435363738394041424344454647484950515253545556575859606162123456789101112131415161718192223242526272829303132333435363738394041424344454647484950515253545556575859606162

最后，编译这个例子，可以参考下面通用的编译命令：

g++ -std=c++11 surf.cpp `pkg-config --cflags opencv` `pkg-config --libs opencv`11

更简化的编译命令：

g++ -std=c++11 surf.cpp -lopencv_xfeatures2d -lopencv_cudafeatures2d