Subversion Repositories OpenCV2-Cookbook

Rev

View as "text/plain" | Blame | Last modification | View Log | RSS feed

1
/*------------------------------------------------------------------------------------------*\
   This file contains material supporting chapter 9 of the cookbook:  
   Computer Vision Programming using the OpenCV Library.
   by Robert Laganiere, Packt Publishing, 2011.

   This program is free software; permission is hereby granted to use, copy, modify,
   and distribute this source code, or portions thereof, for any purpose, without fee,
   subject to the restriction that the copyright notice may not be removed
   or altered from any source or altered source distribution.
   The software is released on an as-is basis and without any warranties of any kind.
   In particular, the software is not guaranteed to be fault-tolerant or free from failure.
   The author disclaims all warranties with regard to this software, any use,
   and any consequent failure, is purely the responsibility of the user.
 
   Copyright (C) 2010-2011 Robert Laganiere, www.laganiere.name
\*------------------------------------------------------------------------------------------*/


#include <iostream>
#include <vector>
#include <opencv2/core/core.hpp>
#include <opencv2/imgproc/imgproc.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <opencv2/features2d/features2d.hpp>
#include <opencv2/calib3d/calib3d.hpp>

int main()
{
        // Read input images
        cv::Mat image1= cv::imread("../church01.jpg",0);
        cv::Mat image2= cv::imread("../church03.jpg",0);
        if (!image1.data || !image2.data)
                return 0;

    // Display the images
        cv::namedWindow("Right Image");
        cv::imshow("Right Image",image1);
        cv::namedWindow("Left Image");
        cv::imshow("Left Image",image2);

        // vector of keypoints
        std::vector<cv::KeyPoint> keypoints1;
        std::vector<cv::KeyPoint> keypoints2;

        // Construction of the SURF feature detector
        cv::SurfFeatureDetector surf(3000);

        // Detection of the SURF features
        surf.detect(image1,keypoints1);
        surf.detect(image2,keypoints2);

        std::cout << "Number of SURF points (1): " << keypoints1.size() << std::endl;
        std::cout << "Number of SURF points (2): " << keypoints2.size() << std::endl;
       
        // Draw the kepoints
        cv::Mat imageKP;
        cv::drawKeypoints(image1,keypoints1,imageKP,cv::Scalar(255,255,255),cv::DrawMatchesFlags::DRAW_RICH_KEYPOINTS);
        cv::namedWindow("Right SURF Features");
        cv::imshow("Right SURF Features",imageKP);
        cv::drawKeypoints(image2,keypoints2,imageKP,cv::Scalar(255,255,255),cv::DrawMatchesFlags::DRAW_RICH_KEYPOINTS);
        cv::namedWindow("Left SURF Features");
        cv::imshow("Left SURF Features",imageKP);

        // Construction of the SURF descriptor extractor
        cv::SurfDescriptorExtractor surfDesc;

        // Extraction of the SURF descriptors
        cv::Mat descriptors1, descriptors2;
        surfDesc.compute(image1,keypoints1,descriptors1);
        surfDesc.compute(image2,keypoints2,descriptors2);

        std::cout << "descriptor matrix size: " << descriptors1.rows << " by " << descriptors1.cols << std::endl;

        // Construction of the matcher
        cv::BruteForceMatcher<cv::L2<float>> matcher;

        // Match the two image descriptors
        std::vector<cv::DMatch> matches;
        matcher.match(descriptors1,descriptors2, matches);

        std::cout << "Number of matched points: " << matches.size() << std::endl;

        // Select few Matches  
        std::vector<cv::DMatch> selMatches;
        /*
        keypoints1.push_back(cv::KeyPoint(342.,615.,2));  
        keypoints2.push_back(cv::KeyPoint(410.,600.,2));
        selMatches.push_back(cv::DMatch(keypoints1.size()-1,keypoints2.size()-1,0)); // street light bulb
        selMatches.push_back(matches[6]);  // right tower
        selMatches.push_back(matches[60]);  // left bottom window
        selMatches.push_back(matches[139]);
        selMatches.push_back(matches[141]);  // middle window
        selMatches.push_back(matches[213]);
        selMatches.push_back(matches[273]);
       
        int kk=0;
        while (kk<matches.size()) {
                std::cout<<kk<<std::endl;
        selMatches.push_back(matches[kk++]);
        selMatches.pop_back();
        cv::waitKey();
        }
        */


        /* between church01 and church03 */
        selMatches.push_back(matches[14]);  
        selMatches.push_back(matches[16]);
        selMatches.push_back(matches[141]);  
        selMatches.push_back(matches[146]);
        selMatches.push_back(matches[235]);
        selMatches.push_back(matches[238]);
        selMatches.push_back(matches[274]);

        // Draw the selected matches
        cv::Mat imageMatches;
        cv::drawMatches(image1,keypoints1,  // 1st image and its keypoints
                            image2,keypoints2,  // 2nd image and its keypoints
//                                      selMatches,                     // the matches
                                        matches,                        // the matches
                                        imageMatches,           // the image produced
                                        cv::Scalar(255,255,255)); // color of the lines
        cv::namedWindow("Matches");
        cv::imshow("Matches",imageMatches);

        // Convert 1 vector of keypoints into
        // 2 vectors of Point2f
        std::vector<int> pointIndexes1;
        std::vector<int> pointIndexes2;
        for (std::vector<cv::DMatch>::const_iterator it= selMatches.begin();
                 it!= selMatches.end(); ++it) {

                         // Get the indexes of the selected matched keypoints
                         pointIndexes1.push_back(it->queryIdx);
                         pointIndexes2.push_back(it->trainIdx);
        }
                 
        // Convert keypoints into Point2f
        std::vector<cv::Point2f> selPoints1, selPoints2;
        cv::KeyPoint::convert(keypoints1,selPoints1,pointIndexes1);
        cv::KeyPoint::convert(keypoints2,selPoints2,pointIndexes2);

        // check by drawing the points
        std::vector<cv::Point2f>::const_iterator it= selPoints1.begin();
        while (it!=selPoints1.end()) {

                // draw a circle at each corner location
                cv::circle(image1,*it,3,cv::Scalar(255,255,255),2);
                ++it;
        }

        it= selPoints2.begin();
        while (it!=selPoints2.end()) {

                // draw a circle at each corner location
                cv::circle(image2,*it,3,cv::Scalar(255,255,255),2);
                ++it;
        }

        // Compute F matrix from 7 matches
        cv::Mat fundemental= cv::findFundamentalMat(
                cv::Mat(selPoints1), // points in first image
                cv::Mat(selPoints2), // points in second image
                CV_FM_7POINT);       // 7-point method

        std::cout << "F-Matrix size= " << fundemental.rows << "," << fundemental.cols << std::endl;  

        // draw the left points corresponding epipolar lines in right image
        std::vector<cv::Vec3f> lines1;
        cv::computeCorrespondEpilines(
                cv::Mat(selPoints1), // image points
                1,                   // in image 1 (can also be 2)
                fundemental, // F matrix
                lines1);     // vector of epipolar lines

        // for all epipolar lines
        for (vector<cv::Vec3f>::const_iterator it= lines1.begin();
                 it!=lines1.end(); ++it) {

                         // draw the epipolar line between first and last column
                         cv::line(image2,cv::Point(0,-(*it)[2]/(*it)[1]),
                                             cv::Point(image2.cols,-((*it)[2]+(*it)[0]*image2.cols)/(*it)[1]),
                                                         cv::Scalar(255,255,255));
        }
               
        // draw the left points corresponding epipolar lines in left image
        std::vector<cv::Vec3f> lines2;
        cv::computeCorrespondEpilines(cv::Mat(selPoints2),2,fundemental,lines2);
        for (vector<cv::Vec3f>::const_iterator it= lines2.begin();
                 it!=lines2.end(); ++it) {

                         // draw the epipolar line between first and last column
                         cv::line(image1,cv::Point(0,-(*it)[2]/(*it)[1]),
                                             cv::Point(image1.cols,-((*it)[2]+(*it)[0]*image1.cols)/(*it)[1]),
                                                         cv::Scalar(255,255,255));
        }
               
    // Display the images with points and epipolar lines
        cv::namedWindow("Right Image Epilines");
        cv::imshow("Right Image Epilines",image1);
        cv::namedWindow("Left Image Epilines");
        cv::imshow("Left Image Epilines",image2);

        /*
        std::nth_element(matches.begin(),    // initial position
                             matches.begin()+matches.size()/2, // 50%
                                         matches.end());     // end position
        // remove all elements after the
        matches.erase(matches.begin()+matches.size()/2, matches.end());
*/

        // Convert keypoints into Point2f
        std::vector<cv::Point2f> points1, points2;
        for (std::vector<cv::DMatch>::const_iterator it= matches.begin();
                 it!= matches.end(); ++it) {

                         // Get the position of left keypoints
                         float x= keypoints1[it->queryIdx].pt.x;
                         float y= keypoints1[it->queryIdx].pt.y;
                         points1.push_back(cv::Point2f(x,y));
                         // Get the position of right keypoints
                         x= keypoints2[it->trainIdx].pt.x;
                         y= keypoints2[it->trainIdx].pt.y;
                         points2.push_back(cv::Point2f(x,y));
        }

        std::cout << points1.size() << " " << points2.size() << std::endl;

        // Compute F matrix using RANSAC
        std::vector<uchar> inliers(points1.size(),0);
        fundemental= cv::findFundamentalMat(
                cv::Mat(points1),cv::Mat(points2), // matching points
                inliers,      // match status (inlier ou outlier)  
                CV_FM_RANSAC, // RANSAC method
                1,            // distance to epipolar line
                0.98);        // confidence probability

        // Read input images
        image1= cv::imread("../church01.jpg",0);
        image2= cv::imread("../church03.jpg",0);

        // Draw the epipolar line of few points
        cv::computeCorrespondEpilines(cv::Mat(selPoints1),1,fundemental,lines1);
        for (vector<cv::Vec3f>::const_iterator it= lines1.begin();
                 it!=lines1.end(); ++it) {

                         cv::line(image2,cv::Point(0,-(*it)[2]/(*it)[1]),
                                             cv::Point(image2.cols,-((*it)[2]+(*it)[0]*image2.cols)/(*it)[1]),
                                                         cv::Scalar(255,255,255));
        }

        cv::computeCorrespondEpilines(cv::Mat(selPoints2),2,fundemental,lines2);
        for (vector<cv::Vec3f>::const_iterator it= lines2.begin();
                 it!=lines2.end(); ++it) {

                         cv::line(image1,cv::Point(0,-(*it)[2]/(*it)[1]),
                                             cv::Point(image1.cols,-((*it)[2]+(*it)[0]*image1.cols)/(*it)[1]),
                                                         cv::Scalar(255,255,255));
        }

        // Draw the inlier points
        std::vector<cv::Point2f> points1In, points2In;
        std::vector<cv::Point2f>::const_iterator itPts= points1.begin();
        std::vector<uchar>::const_iterator itIn= inliers.begin();
        while (itPts!=points1.end()) {

                // draw a circle at each inlier location
                if (*itIn) {
                        cv::circle(image1,*itPts,3,cv::Scalar(255,255,255),2);
                        points1In.push_back(*itPts);
                }
                ++itPts;
                ++itIn;
        }

        itPts= points2.begin();
        itIn= inliers.begin();
        while (itPts!=points2.end()) {

                // draw a circle at each inlier location
                if (*itIn) {
                        cv::circle(image2,*itPts,3,cv::Scalar(255,255,255),2);
                        points2In.push_back(*itPts);
                }
                ++itPts;
                ++itIn;
        }

    // Display the images with points
        cv::namedWindow("Right Image Epilines (RANSAC)");
        cv::imshow("Right Image Epilines (RANSAC)",image1);
        cv::namedWindow("Left Image Epilines (RANSAC)");
        cv::imshow("Left Image Epilines (RANSAC)",image2);

        cv::findHomography(cv::Mat(points1In),cv::Mat(points2In),inliers,CV_RANSAC,1.);

        // Read input images
        image1= cv::imread("../church01.jpg",0);
        image2= cv::imread("../church03.jpg",0);

        // Draw the inlier points
        itPts= points1In.begin();
        itIn= inliers.begin();
        while (itPts!=points1In.end()) {

                // draw a circle at each inlier location
                if (*itIn)
                        cv::circle(image1,*itPts,3,cv::Scalar(255,255,255),2);
               
                ++itPts;
                ++itIn;
        }

        itPts= points2In.begin();
        itIn= inliers.begin();
        while (itPts!=points2In.end()) {

                // draw a circle at each inlier location
                if (*itIn)
                        cv::circle(image2,*itPts,3,cv::Scalar(255,255,255),2);
               
                ++itPts;
                ++itIn;
        }

    // Display the images with points
        cv::namedWindow("Right Image Homography (RANSAC)");
        cv::imshow("Right Image Homography (RANSAC)",image1);
        cv::namedWindow("Left Image Homography (RANSAC)");
        cv::imshow("Left Image Homography (RANSAC)",image2);

        cv::waitKey();
        return 0;
        }