《视觉slam十四讲》初学小白笔记（7）

提取ORB特征点进行特征匹配 

高翔大佬的代码用的是opencv3，所以下面两句有点不一样~如果使用opencv2，则是create("ORB"),否则报错。

Ptr<FeatureDetector> detector =ORB::create();	//创建ORB特征点检测
Ptr<DescriptorExtractor> descriptor =ORB::create();	//使用ORB特征来描述特征点

#include <iostream>
#include <opencv2/core/core.hpp>
#include <opencv2/features2d/features2d.hpp>
#include <opencv2/highgui/highgui.hpp>

using namespace std;
using namespace cv;

int main(int argc,char** argv){
    if(argc!=3){
      cout<<"usage:feature_extraction img1，img2"<<endl;
      return -1;
    }
    
    //读取图像
    Mat img_1=imread(argv[1],CV_LOAD_IMAGE_COLOR);
    Mat img_2=imread(argv[2],CV_LOAD_IMAGE_COLOR);
    
    //初始化
    std::vector<KeyPoint> keypoints_1,keypoints_2;	//创建关键点
    Mat descriptors_1,descriptors_2;			//创建描述子
    
    //opencv2中create()要加内容
    Ptr<FeatureDetector> detector =ORB::create("ORB");	//创建ORB特征点检测
    Ptr<DescriptorExtractor> descriptor =ORB::create("ORB");	//使用ORB特征来描述特征点
    Ptr<DescriptorMatcher> matcher=DescriptorMatcher::create("BruteForce-Hamming");//设置特征匹配使用的汉明距离计算

    //设置特征点检测
    //Ptr<FeatureDetector> detector = FeatureDetector::create(dectector_name);
    //Ptr<DescriptorExtractor> descriptor = DescriptorExtractor::create(descriptor_name);
    
    //第一步：检测oriented FAST角点检测
    detector->detect(img_1,keypoints_1);
    detector->detect(img_2,keypoints_2);
    
    //第二步：根据角点位置计算BRIEF描述子
    descriptor->compute(img_1,keypoints_1,descriptors_1);
    descriptor->compute(img_2,keypoints_2,descriptors_2);
    
    Mat outimg1;
    drawKeypoints(img_1,keypoints_1,outimg1,Scalar::all(-1),DrawMatchesFlags::DEFAULT);
    imshow("ORB feature points",outimg1);
    waitKey(0);
    
    //第三步：对两幅图像中的BRIEF描述子进行匹配，使用hamming距离
    vector<DMatch> matches;
    matcher->match(descriptors_1,descriptors_2,matches);
    
    //第四步：绘制所有匹配结果
    Mat img_match;
    drawMatches(img_1,keypoints_1,img_2,keypoints_2,matches,img_match);
    imshow("All match point pairs",img_match);
    waitKey(0);
    
    //第五步：匹配点对筛选
    double min_dist=10000,max_dist=0;//最小10000，最大0，通过反向计算求得max和min
    
    //找出匹配点之间的最小距离和最大距离
    for(int i=0;i<descriptors_1.rows;i++){
      double dist=matches[i].distance;
      if(dist<min_dist) min_dist=dist;
      if(dist>max_dist) max_dist=dist;
    }
    
    // 仅供娱乐的写法
    //min_dist = min_element( matches.begin(), matches.end(), [](const DMatch& m1, const DMatch& m2) {return m1.distance<m2.distance;} )->distance;
    //max_dist = max_element( matches.begin(), matches.end(), [](const DMatch& m1, const DMatch& m2) {return m1.distance<m2.distance;} )->distance;

    //显示所有匹配之间的最大距离和最小距离
    printf("Max dist:%f\n",max_dist);
    printf("min_dist:%f\n",min_dist);
    
    //当描述子之间的距离大于两倍的最小距离是，即认为匹配错误
    //但有时候最小距离会非常小，设置一个经验值30作为下限
    std::vector<DMatch>good_matches;
    for(int i=0;i<descriptors_1.rows;i++){
      if(matches[i].distance<=max(2*min_dist,30.0))	//取两者较大值
	good_matches.push_back(matches[i]);
    }
    
    //第六步：绘制筛选后的匹配结果
    Mat img_goodMatch;
    drawMatches(img_1,keypoints_1,img_2,keypoints_2,good_matches,img_goodMatch);
    imshow("Match point pairs after optimization",img_goodMatch);
    waitKey(0);
    
    return 0;
}

CMakeLists.txt

cmake_minimum_required( VERSION 2.8 )
project( vo1 )

set( CMAKE_BUILD_TYPE "Release" )
set( CMAKE_CXX_FLAGS "-std=c++11 -O3" )

#opencv2
find_package( OpenCV REQUIRED )

include_directories( 
    ${OpenCV_INCLUDE_DIRS} 
    ${CSPARSE_INCLUDE_DIR}
    "/usr/include/eigen3/"
)

add_executable( feature_extraction feature_extraction.cc  )
target_link_libraries( feature_extraction ${OpenCV_LIBS} )

 

运行结果：

ORB特征点图

 显示所有匹配点

筛选过后的匹配点