SIFT配合暴力匹配进行关键点描述和提取-阿里云开发者社区

SIFT配合暴力匹配进行关键点描述和提取

2022-06-26 202

版权

本文内容由阿里云实名注册用户自发贡献，版权归原作者所有，阿里云开发者社区不拥有其著作权，亦不承担相应法律责任。具体规则请查看《阿里云开发者社区用户服务协议》和《阿里云开发者社区知识产权保护指引》。如果您发现本社区中有涉嫌抄袭的内容，填写侵权投诉表单进行举报，一经查实，本社区将立刻删除涉嫌侵权内容。

简介： SIFT配合暴力匹配进行关键点描述和提取

接上一篇图片匹配教程

此处为语雀内容卡片，点击链接查看：https://developer.aliyun.com/article/978422?spm=a2c6h.13148508.setting.20.20cb4f0e3oNmeo

和上次ORB对比一下

上次的效果

本次的效果

上次的效果

本次的效果

可以说比上次好的多得多

测试环境

win10

visual studio 2019

opencv-3.4.3

opencv_contrib-3.4.3

cmake 3.20.3

备注

2021年3月6日，已满20年，专利权已经到期！SIFT已经成为全人类的公共技术，任何人和组织都可以免费使用！

opencv4.5.2已将sift移入主库

sift是有专利的算法, cmake编译时需要勾选 OPENCV_ENABLE_NONFREE, 才能使用有专利的算法

opencv源码编译可参考该视频教程:

vs2019+opencv3.4.6+contrib3.4.6源码编译

测试源码

#include<iostream>
#include<opencv2/opencv.hpp>
#include<opencv2/features2d/features2d.hpp>
#include<opencv2/xfeatures2d/nonfree.hpp>
using namespace std;
using namespace cv;
int main()
{
  Mat trainImage = imread("D:/script/myGit/autojs-flann-feature-matching/图片素材/game001.png"), trainImage_gray;
  //imshow("trainImage", trainImage);
  cvtColor(trainImage, trainImage_gray, COLOR_BGR2GRAY);
  vector<KeyPoint>train_keyPoint;
  Mat trainDescription;
  //SiftFeatureDetector featureDetector;
  //featureDetector.detect(trainImage_gray, train_keyPoint);
  Ptr<xfeatures2d::SiftFeatureDetector> featureDetector = xfeatures2d::SiftFeatureDetector::create();
  featureDetector->detect(trainImage_gray, train_keyPoint);
  Ptr<xfeatures2d::SiftDescriptorExtractor> featureExtractor = xfeatures2d::SiftDescriptorExtractor::create();
  featureExtractor->compute(trainImage_gray, train_keyPoint, trainDescription);
  BFMatcher matcher;
  vector<Mat>train_desc_collection(1, trainDescription);
  matcher.add(train_desc_collection);
  matcher.train();
  Mat captureImage, captureImage_gray;
  captureImage = imread("D:/script/myGit/autojs-flann-feature-matching/图片素材/game001_女生头像.png");
  cvtColor(captureImage, captureImage_gray, COLOR_BGR2GRAY);
  vector<KeyPoint>test_keyPoint;
  Mat testDescriptor;
  featureDetector->detect(captureImage_gray, test_keyPoint);
  featureExtractor->compute(captureImage_gray, test_keyPoint, testDescriptor);
  vector<vector<DMatch>>matches;
  matcher.knnMatch(testDescriptor, matches, 2);
  vector<DMatch>goodMatches;
  for (unsigned int i = 0; i < matches.size(); i++)
  {
    if (matches[i][0].distance < 0.6 * matches[i][1].distance)
      goodMatches.push_back(matches[i][0]);
  }
  Mat dstImage;
  drawMatches(captureImage, test_keyPoint, trainImage, train_keyPoint, goodMatches, dstImage);
  imshow("dstImage", dstImage);
  cv::waitKey(0);
  return 0;
}