I'm using SURF algorithm for real time hand detection.I'll let the user put his hand in a rectangle and then use it as my object and find the hand in first frame using that object,in the next loop i'll set the detected hand as my object and so on...
But when running it shows the following error
OpenCV Error: Assertion failed (count >= 4) in cvFindHomography, file /build/buildd/opencv-2.4.8+dfsg1/modules/calib3d/src/fundam.cpp, line 235
terminate called after throwing an instance of 'cv::Exception'
what(): /build/buildd/opencv-2.4.8+dfsg1/modules/calib3d/src/fundam.cpp:235: error: (-215) count >= 4 in function cvFindHomography
Here's my code:
#include <stdio.h>
#include <iostream>
#include "opencv2/imgproc/imgproc.hpp"
#include "opencv2/core/core.hpp"
#include "opencv2/features2d/features2d.hpp"
#include "opencv2/highgui/highgui.hpp"
#include "opencv2/calib3d/calib3d.hpp"
#include "opencv2/nonfree/nonfree.hpp"
using namespace cv;
using namespace std;
int main()
{
char k;
Mat img_object,img_scene,frame;
VideoCapture cap(0);
if(cap.isOpened()==0)
{
cout<<"ERROR";
return -1;
}
while(1)
{
cap>>frame;
rectangle(frame,Point(100,100),Point(300,300),(0,255,0),4,8,0);
imshow("gig",frame);
k=waitKey(1);
if(k=='q')
{
Mat img_object1(frame,Rect(25,25,100,100));
img_object=img_object1.clone();
cvtColor(img_object,img_object,CV_BGR2GRAY);
break;
}
}
while(1)
{
cap>>img_scene;
cvtColor(img_scene,img_scene,CV_BGR2GRAY);
int minHessian = 50;
SurfFeatureDetector detector( minHessian );
std::vector<KeyPoint> keypoints_object, keypoints_scene;
detector.detect( img_object, keypoints_object );
detector.detect( img_scene, keypoints_scene );
SurfDescriptorExtractor extractor;
Mat descriptors_object, descriptors_scene;
extractor.compute( img_object, keypoints_object, descriptors_object );
extractor.compute( img_scene, keypoints_scene, descriptors_scene );
FlannBasedMatcher matcher;
std::vector< DMatch > matches;
matcher.match( descriptors_object, descriptors_scene, matches );
double max_dist = 0; double min_dist = 100;
for( int i = 0; i < descriptors_object.rows; i++ )
{
double dist = matches[i].distance;
if( dist < min_dist ) min_dist = dist;
if( dist > max_dist ) max_dist = dist;
}
printf("-- Max dist : %f \n", max_dist );
printf("-- Min dist : %f \n", min_dist );
std::vector< DMatch > good_matches;
for( int i = 0; i < descriptors_object.rows; i++ )
{
if( matches[i].distance < 3*min_dist )
{ good_matches.push_back( matches[i]); }
}
Mat img_matches;
drawMatches( img_object, keypoints_object, img_scene, keypoints_scene,
good_matches, img_matches, Scalar::all(-1), Scalar::all(-1),
vector<char>(), DrawMatchesFlags::NOT_DRAW_SINGLE_POINTS );
std::vector<Point2f> obj;
std::vector<Point2f> scene;
for( int i = 0; i < good_matches.size(); i++ )
{
obj.push_back( keypoints_object[ good_matches[i].queryIdx ].pt );
scene.push_back( keypoints_scene[ good_matches[i].trainIdx ].pt );
}
Mat H = findHomography( obj, scene, CV_RANSAC );
std::vector<Point2f> obj_corners(4);
obj_corners[0] = cvPoint(0,0); obj_corners[1] = cvPoint( img_object.cols, 0 );
obj_corners[2] = cvPoint( img_object.cols, img_object.rows ); obj_corners[3] = cvPoint( 0, img_object.rows );
std::vector<Point2f> scene_corners(4);
perspectiveTransform( obj_corners, scene_corners, H);
line( img_matches, scene_corners[0] + Point2f( img_object.cols, 0), scene_corners[1] + Point2f( img_object.cols, 0), Scalar(0, 255, 0), 4 );
line( img_matches, scene_corners[1] + Point2f( img_object.cols, 0), scene_corners[2] + Point2f( img_object.cols, 0), Scalar( 0, 255, 0), 4 );
line( img_matches, scene_corners[2] + Point2f( img_object.cols, 0), scene_corners[3] + Point2f( img_object.cols, 0), Scalar( 0, 255, 0), 4 );
line( img_matches, scene_corners[3] + Point2f( img_object.cols, 0), scene_corners[0] + Point2f( img_object.cols, 0), Scalar( 0, 255, 0), 4 );
Mat nn(img_scene,Rect((scene_corners[0] ).x,(scene_corners[0] ).y,(scene_corners[2] ).x,(scene_corners[2]).y));
imshow( "Good Matches & Object detection", img_matches );
img_object=nn.clone();
k=waitKey(1);
if(k=='q')
{
break;
}
}
return 0;
}
Related
I'm new to OpenCV platform. I've installed and successfully run some tutorial codes(inbuilt). But I'm facing some issues with the codes involving haarcascade.The code builds perfectly and also the webcam seems to flash light and work but the code displays blank window as attached.image
Can anyone suggest some rectification for the problem?
(I'm using opencv version 3.0.0 and Visual Studios 2013.)
#include "stdafx.h"
#include "opencv2/objdetect.hpp"
#include "opencv2/videoio.hpp"
#include "opencv2/highgui.hpp"
#include "opencv2/imgproc.hpp"
#include <iostream>
#include <stdio.h>
using namespace std;
using namespace cv;
/** Function Headers */
void detectAndDisplay( Mat frame );
/** Global variables */
String face_cascade_name = "haarcascade_frontalface_alt.xml";
String eyes_cascade_name = "haarcascade_eye_tree_eyeglasses.xml";
CascadeClassifier face_cascade;
CascadeClassifier eyes_cascade;
String window_name = "Capture - Face detection";
/** #function main */
int main( void )
{
VideoCapture capture;
Mat frame;
//-- 1. Load the cascades
if( !face_cascade.load( face_cascade_name ) ){ printf("--(!)Error loading face cascade\n"); return -1; };
if( !eyes_cascade.load( eyes_cascade_name ) ){ printf("--(!)Error loading eyes cascade\n"); return -1; };
//-- 2. Read the video stream
capture.open( -1 );
if ( ! capture.isOpened() ) { printf("--(!)Error opening video capture\n"); return -1; }
while ( capture.read(frame) )
{
if( frame.empty() )
{
printf(" --(!) No captured frame -- Break!");
break;
}
//-- 3. Apply the classifier to the frame
detectAndDisplay( frame );
int c = waitKey(10);
if( (char)c == 27 ) { break; } // escape
}
return 0;
}
/** #function detectAndDisplay */
void detectAndDisplay( Mat frame )
{
std::vector<Rect> faces;
Mat frame_gray;
cvtColor( frame, frame_gray, COLOR_BGR2GRAY );
equalizeHist( frame_gray, frame_gray );
//-- Detect faces
face_cascade.detectMultiScale( frame_gray, faces, 1.1, 2, 0|CASCADE_SCALE_IMAGE, Size(30, 30) );
for ( size_t i = 0; i < faces.size(); i++ )
{
Point center( faces[i].x + faces[i].width/2, faces[i].y + faces[i].height/2 );
ellipse( frame, center, Size( faces[i].width/2, faces[i].height/2 ), 0, 0, 360, Scalar( 255, 0, 255 ), 4, 8, 0 );
Mat faceROI = frame_gray( faces[i] );
std::vector<Rect> eyes;
//-- In each face, detect eyes
eyes_cascade.detectMultiScale( faceROI, eyes, 1.1, 2, 0 |CASCADE_SCALE_IMAGE, Size(30, 30) );
for ( size_t j = 0; j < eyes.size(); j++ )
{
Point eye_center( faces[i].x + eyes[j].x + eyes[j].width/2, faces[i].y + eyes[j].y + eyes[j].height/2 );
int radius = cvRound( (eyes[j].width + eyes[j].height)*0.25 );
circle( frame, eye_center, radius, Scalar( 255, 0, 0 ), 4, 8, 0 );
}
}
//-- Show what you got
imshow( window_name, frame );
}
i think it is a known bug
could you try my updated code ( i pointed out the changes by comments)
#include "opencv2/objdetect.hpp"
#include "opencv2/videoio.hpp"
#include "opencv2/highgui.hpp"
#include "opencv2/imgproc.hpp"
#include <opencv2/core/ocl.hpp> // additional header ************************
#include <iostream>
#include <stdio.h>
using namespace std;
using namespace cv;
/** Function Headers */
void detectAndDisplay( Mat frame );
/** Global variables */
String face_cascade_name = "haarcascade_frontalface_alt.xml";
String eyes_cascade_name = "haarcascade_eye_tree_eyeglasses.xml";
CascadeClassifier face_cascade;
CascadeClassifier eyes_cascade;
String window_name = "Capture - Face detection";
/** #function main */
int main( void )
{
ocl::setUseOpenCL(false); // disable OpenCL *******************
VideoCapture capture;
Mat frame;
//-- 1. Load the cascades
if( !face_cascade.load( face_cascade_name ) ){ printf("--(!)Error loading face cascade\n"); return -1; };
if( !eyes_cascade.load( eyes_cascade_name ) ){ printf("--(!)Error loading eyes cascade\n"); return -1; };
//-- 2. Read the video stream
capture.open( 0 ); // Open webcam 0 ***************************
if ( ! capture.isOpened() ) { printf("--(!)Error opening video capture\n"); return -1; }
while ( capture.read(frame) )
{
if( frame.empty() )
{
printf(" --(!) No captured frame -- Break!");
break;
}
//-- 3. Apply the classifier to the frame
detectAndDisplay( frame );
int c = waitKey(10);
if( (char)c == 27 ) { break; } // escape
}
return 0;
}
/** #function detectAndDisplay */
void detectAndDisplay( Mat frame )
{
std::vector<Rect> faces;
Mat frame_gray;
cvtColor( frame, frame_gray, COLOR_BGR2GRAY );
equalizeHist( frame_gray, frame_gray );
//-- Detect faces
face_cascade.detectMultiScale( frame_gray, faces, 1.1, 2, 0|CASCADE_SCALE_IMAGE, Size(30, 30) );
for ( size_t i = 0; i < faces.size(); i++ )
{
Point center( faces[i].x + faces[i].width/2, faces[i].y + faces[i].height/2 );
ellipse( frame, center, Size( faces[i].width/2, faces[i].height/2 ), 0, 0, 360, Scalar( 255, 0, 255 ), 4, 8, 0 );
Mat faceROI = frame_gray( faces[i] );
std::vector<Rect> eyes;
//-- In each face, detect eyes
eyes_cascade.detectMultiScale( faceROI, eyes, 1.1, 2, 0 |CASCADE_SCALE_IMAGE, Size(30, 30) );
for ( size_t j = 0; j < eyes.size(); j++ )
{
Point eye_center( faces[i].x + eyes[j].x + eyes[j].width/2, faces[i].y + eyes[j].y + eyes[j].height/2 );
int radius = cvRound( (eyes[j].width + eyes[j].height)*0.25 );
circle( frame, eye_center, radius, Scalar( 255, 0, 0 ), 4, 8, 0 );
}
}
//-- Show what you got
imshow( window_name, frame );
}
I try to match multi-object with rotation using a simple template like a smile face template
,and I wanna detect it in the test image like test image
I have tried to using Features2D and Homography to detect, however there are many problems.
P1: It seems this keypoints matching method is not accurate for SIMPLE template(I have tried this method in another template which is much more complicated, the matching result is better). Is there any method on this problem?
P2: Definitely this method is not suitable in multi-object test image. How could I match multiple objects using a single template?(the premise is I don't know the number and location of objects in the template)
Below is my function code.
`//load image
Mat img1 = imread( "2.png", CV_LOAD_IMAGE_GRAYSCALE );
Mat img2 = imread( "1.png", CV_LOAD_IMAGE_GRAYSCALE );
//-- Step 1: Detect the keypoints using SURF Detector
SurfFeatureDetector detector( hessian );
vector<KeyPoint> keypoints1, keypoints2;
detector.detect( img1, keypoints1 );
detector.detect( img2, keypoints2 );
//-- Step 2: Extract the keypoints using SURF Extractor
Mat descriptors1,descriptors2;// extract keypoints
SurfDescriptorExtractor extractor; //Create Descriptor Extractor
extractor.compute( img1, keypoints1, descriptors1 );
extractor.compute( img2, keypoints2, descriptors2 );
//-- Step 3: Matching descriptor vectors using FLANN matcher
FlannBasedMatcher matcher;
std::vector< DMatch > matches;
matcher.match( descriptors_object, descriptors_scene, matches );
double max_dist = 0; double min_dist = 100;
//-- Quick calculation of max and min distances between keypoints
for( int i = 0; i < descriptors_object.rows; i++ )
{ double dist = matches[i].distance;
if( dist < min_dist ) min_dist = dist;
if( dist > max_dist ) max_dist = dist;
}
//-- Draw only "good" matches
std::vector< DMatch > good_matches;
for( int i = 0; i < descriptors_object.rows; i++ )
{ if( matches[i].distance < 3*min_dist )
{ good_matches.push_back( matches[i]); }
}
Mat img_matches;
drawMatches( img_object, keypoints_object, img_scene, keypoints_scene,
good_matches, img_matches, Scalar::all(-1), Scalar::all(-1),
vector<char>(), DrawMatchesFlags::NOT_DRAW_SINGLE_POINTS );
//-- Localize the object
std::vector<Point2f> obj;
std::vector<Point2f> scene;
for( int i = 0; i < good_matches.size(); i++ )
{
//-- Get the keypoints from the good matches
obj.push_back( keypoints_object[ good_matches[i].queryIdx ].pt );
scene.push_back( keypoints_scene[ good_matches[i].trainIdx ].pt );
}
Mat H = findHomography( obj, scene, CV_RANSAC );
//-- Get the corners from the image_1 ( the object to be "detected" )
std::vector<Point2f> obj_corners(4);
obj_corners[0] = cvPoint(0,0); obj_corners[1] = cvPoint( img_object.cols,0 );
obj_corners[2] = cvPoint( img_object.cols, img_object.rows ); obj_corners[3] = cvPoint( 0, img_object.rows );
std::vector<Point2f> scene_corners(4);
perspectiveTransform( obj_corners, scene_corners, H);
//-- Draw lines between the corners (the mapped object in the scene - image_2 )
line( img_matches, scene_corners[0] + Point2f( img_object.cols, 0), scene_corners[1] + Point2f( img_object.cols, 0), Scalar(0, 255, 0), 4 );
line( img_matches, scene_corners[1] + Point2f( img_object.cols, 0), scene_corners[2] + Point2f( img_object.cols, 0), Scalar( 0, 255, 0), 4 );
line( img_matches, scene_corners[2] + Point2f( img_object.cols, 0), scene_corners[3] + Point2f( img_object.cols, 0), Scalar( 0, 255, 0), 4 );
line( img_matches, scene_corners[3] + Point2f( img_object.cols, 0), scene_corners[0] + Point2f( img_object.cols, 0), Scalar( 0, 255, 0), 4 );
`
I am a beginner in computer-vision,and it is my first time asking on this forum. Many thanks for your help!
If your problem is to detect only that kind of images, a simple thing that you can do is to use a circle detector. And you can group the point of the bigger circle (head) and the points of the eyes. If you know the position of the centroids of those 3 circles, you can have the position and rotation of the face by studying where are the eyes.
In the image, the red points represent the centroids of the circles, you can get the head position by finding where the main centroid is, alpha is the angle between the right eye and the main centroid. If you can find the new angle you can compute theta which will indicate the rotation of the face, and maybe this could work even scale changes
I am trying to detect an object using the SurfFeatureDetect and FLANN matcher. However, the code is not able to detect the image accurately. I have also posted the results in pictorial format.
Here's my code from the opencv tutorial website
int main(int argc, char** argv){
if (argc != 3){
readme(); return -1;
}
Mat img_object = imread(argv[1], CV_LOAD_IMAGE_GRAYSCALE);
Mat img_scene = imread(argv[2], CV_LOAD_IMAGE_GRAYSCALE);
if (!img_object.data || !img_scene.data)
{
std::cout << " --(!) Error reading images " << std::endl; return -1;
}
//-- Step 1: Detect the keypoints using SURF Detector
int minHessian = 100;
SurfFeatureDetector detector(minHessian);
std::vector<KeyPoint> keypoints_object, keypoints_scene;
detector.detect(img_object, keypoints_object);
detector.detect(img_scene, keypoints_scene);
//-- Step 2: Calculate descriptors (feature vectors)
SurfDescriptorExtractor extractor;
Mat descriptors_object, descriptors_scene;
extractor.compute(img_object, keypoints_object, descriptors_object);
extractor.compute(img_scene, keypoints_scene, descriptors_scene);
//-- Step 3: Matching descriptor vectors using FLANN matcher
FlannBasedMatcher matcher;
std::vector< DMatch > matches;
matcher.match(descriptors_object, descriptors_scene, matches);
double max_dist = 0; double min_dist = 100;
//-- Quick calculation of max and min distances between keypoints
for (int i = 0; i < descriptors_object.rows; i++)
{
double dist = matches[i].distance;
if (dist < min_dist) min_dist = dist;
if (dist > max_dist) max_dist = dist;
}
printf("-- Max dist : %f \n", max_dist);
printf("-- Min dist : %f \n", min_dist);
//-- Draw only "good" matches (i.e. whose distance is less than 3*min_dist )
std::vector< DMatch > good_matches;
for (int i = 0; i < descriptors_object.rows; i++)
{
if (matches[i].distance < 3 * min_dist)
{
good_matches.push_back(matches[i]);
}
}
Mat img_matches;
drawMatches(img_object, keypoints_object, img_scene, keypoints_scene,
good_matches, img_matches, Scalar::all(-1), Scalar::all(-1),
vector<char>(), DrawMatchesFlags::NOT_DRAW_SINGLE_POINTS);
//-- Localize the object
std::vector<Point2f> obj;
std::vector<Point2f> scene;
for (int i = 0; i < good_matches.size(); i++)
{
//-- Get the keypoints from the good matches
obj.push_back(keypoints_object[good_matches[i].queryIdx].pt);
scene.push_back(keypoints_scene[good_matches[i].trainIdx].pt);
}
Mat H = findHomography(obj, scene, CV_RANSAC);
//-- Get the corners from the image_1 ( the object to be "detected" )
std::vector<Point2f> obj_corners(4);
obj_corners[0] = cvPoint(0, 0); obj_corners[1] = cvPoint(img_object.cols, 0);
obj_corners[2] = cvPoint(img_object.cols, img_object.rows); obj_corners[3] = cvPoint(0, img_object.rows);
std::vector<Point2f> scene_corners(4);
perspectiveTransform(obj_corners, scene_corners, H);
//-- Draw lines between the corners (the mapped object in the scene - image_2 )
line(img_matches, scene_corners[0] + Point2f(img_object.cols, 0), scene_corners[1] + Point2f(img_object.cols, 0), Scalar(0, 255, 0), 4);
line(img_matches, scene_corners[1] + Point2f(img_object.cols, 0), scene_corners[2] + Point2f(img_object.cols, 0), Scalar(0, 255, 0), 4);
line(img_matches, scene_corners[2] + Point2f(img_object.cols, 0), scene_corners[3] + Point2f(img_object.cols, 0), Scalar(0, 255, 0), 4);
line(img_matches, scene_corners[3] + Point2f(img_object.cols, 0), scene_corners[0] + Point2f(img_object.cols, 0), Scalar(0, 255, 0), 4);
//-- Show detected matches
imshow("Good Matches & Object detection", img_matches);
waitKey(0);
return 0;}
/** #function readme */
void readme()
{
std::cout << " Usage: ./SURF_descriptor <img1> <img2>" << std::endl;}
That is a very common failure. The problem is that the homography has 8 degree of freedom (8DOF). This means that you need at least 4 correct correspondences to calculate a good homography:
As you can see, the homography has 8 parameters (the last parameter h33 is just a scale factor).
The problem arises when other than good corrspondces (inlier) you need to filter out bad correspondences (outlier). When the are more outliers than inliers (total/outliers > 50%) the RANSAC procedure cannot find the outlier and you obtain weird results.
Solutions to this problem are not easy. You could:
Use a training image with a similar out-of-plane rotation (and a similar scale) of the object in your query image.
Or, use a transformation with less degree of freedom (such as similarity transform). In this way you will need less inliers. Altho OpenCV lacks support for this simpler transformation with a robust fitting method.
Anyone familiar with this error? I tested a surf descriptor in real-time and it worked well but after few seconds it crashes and I got this error.
It was related when no points were detected. I run my code again and have the detected object stays for more than 2 mins and still no error. but when I removed the object and there were no points, it crashes again after 40 secs.
#include <stdio.h>
#include <iostream>
#include <fstream>
#include <string>
#include "opencv2/core/core.hpp"
#include "opencv2/features2d/features2d.hpp"
#include "opencv2/highgui/highgui.hpp"
#include "opencv2/imgproc/imgproc.hpp"
#include "opencv2/calib3d/calib3d.hpp"
#include "opencv2/nonfree/features2d.hpp"
#include "opencv2/legacy/legacy.hpp"
using namespace cv;
using namespace std;
char key = 'a';
int framecount = 0;
SurfFeatureDetector detector(1000);
SurfDescriptorExtractor extractor;
FlannBasedMatcher matcher;
Mat frame, des_object, image;
Mat des_image, img_matches, H;
std::vector<KeyPoint> kp_object;
std::vector<Point2f> obj_corners(4);
std::vector<KeyPoint> kp_image;
std::vector<vector<DMatch > > matches;
std::vector<DMatch > good_matches;
std::vector<Point2f> obj;
std::vector<Point2f> scene;
std::vector<Point2f> scene_corners(4);
int main()
{
//reference image
Mat object = imread("D:/milo.jpg", CV_LOAD_IMAGE_GRAYSCALE );
if( !object.data )
{
std::cout<< "Error reading object " << std::endl;
return -1;
}
//compute detectors and descriptors of reference image
detector.detect( object, kp_object );
extractor.compute( object, kp_object, des_object );
//create video capture object
CvCapture* capture = cvCaptureFromCAM(0);
cvSetCaptureProperty(capture, CV_CAP_PROP_FRAME_WIDTH, 270);
cvSetCaptureProperty(capture, CV_CAP_PROP_FRAME_HEIGHT, 190);
//Get the corners from the object
obj_corners[0] = cvPoint(0,0);
obj_corners[1] = cvPoint( object.cols, 0 );
obj_corners[2] = cvPoint( object.cols, object.rows );
obj_corners[3] = cvPoint( 0, object.rows );
//wile loop for real time detection
while (key != 27)
{
Mat frame;
frame = cvQueryFrame(capture);
if (framecount < 5)
{
framecount++;
continue;
}
Mat des_image, img_matches;
std::vector<KeyPoint> kp_image;
std::vector<vector<DMatch > > matches;
std::vector<DMatch > good_matches;
std::vector<Point2f> obj;
std::vector<Point2f> scene;
std::vector<Point2f> scene_corners(4);
Mat H;
Mat image;
cvtColor(frame, image, CV_RGB2GRAY);
detector.detect( image, kp_image );
extractor.compute( image, kp_image, des_image );
matcher.knnMatch(des_object, des_image, matches, 2);
int goodMatchesCounter =0;
for(int i = 0; i < min(des_image.rows-1,(int) matches.size()); i++) //THIS LOOP IS
SENSITIVE TO SEGFAULTS
{
if(((int)matches[i].size()<=2 && (int)matches[i].size()>0) && (matches[i}
[0].distance<0.6*(matches[i][1].distance)))
{
// good_matches.push_back(matches[i][0]);
obj.push_back( kp_object[ matches[i][0].queryIdx ].pt );
scene.push_back( kp_image[ matches[i][0].trainIdx ].pt );
goodMatchesCounter++;
}
}
//Draw only "good" matches
// drawMatches( object, kp_object, image, kp_image, good_matches, img_matches,
Scalar::all(-1), Scalar::all(-1), vector<char>(),
DrawMatchesFlags::NOT_DRAW_SINGLE_POINTS );
if (goodMatchesCounter >= 4)
{
H = findHomography( obj, scene, CV_RANSAC );
perspectiveTransform( obj_corners, scene_corners, H);
//Draw lines between the corners (the mapped object in the scene image )
line( image, scene_corners[0], scene_corners[1], Scalar( 0, 0, 0), 4 );
line( image, scene_corners[1], scene_corners[2], Scalar( 0, 0, 0),
4 );
line( image, scene_corners[2], scene_corners[3], Scalar( 0, 0, 0),
4 );
line( image, scene_corners[3], scene_corners[0], Scalar( 0, 0, 0),
4 );
}
//Show detected matches
imshow( "Good Matches", image );
key = waitKey(1);
}
return 0;
}
OpenCV 2.4.5, CUDA 5.0
I tried to transfer my SURF matcher from the CPU to the GPU and got such a strange result. I use knnMatch and findHomography + perspectiveTransform together with my function, which checks the corners of the bounding box for the result to more precision.
GPU part:
const int baseImagesSize = baseImages.size();
SURF_GPU surf(1500);
surf.extended = false;
GpuMat keypoints_test_GPU, descriptors_test_GPU;
surf(frame, GpuMat(), keypoints_test_GPU, descriptors_test_GPU);
vector<float> descriptors_test_CPU;
surf.downloadDescriptors(descriptors_test_GPU, descriptors_test_CPU);
Mat descriptors_test_CPU_Mat(descriptors_test_CPU);
vector<Point2f> objs_corners(4);
BruteForceMatcher_GPU< L2<float> > matcher;
vector<KeyPoint> keypoints_test_CPU;
surf.downloadKeypoints(keypoints_test_GPU, keypoints_test_CPU);
for (int i = 0; i < baseImagesSize; ++i)
{
//Get the corners from the object
objs_corners[0] = cvPoint(0,0);
objs_corners[1] = cvPoint( baseImages[i].cols, 0 );
objs_corners[2] = cvPoint( baseImages[i].cols, baseImages[i].rows );
objs_corners[3] = cvPoint( 0, baseImages[i].rows );
//cout<<endl<<objs_corners[0]<<" "<<objs_corners[1]<<" "<<objs_corners[2]<<" "<<objs_corners[3]<<endl;
GpuMat keypoints_tmp_GPU, descriptors_tmp_GPU;
surf(baseImages[i], GpuMat(), keypoints_tmp_GPU, descriptors_tmp_GPU);
GpuMat trainIdx, distance;
vector< vector<DMatch> > matches;
matcher.knnMatch(descriptors_test_GPU, descriptors_tmp_GPU, matches, 2);
vector<KeyPoint> keypoints_tmp_CPU;
surf.downloadKeypoints(keypoints_tmp_GPU, keypoints_tmp_CPU);
std::vector<DMatch > good_matches;
for(int k = 0; k < min(descriptors_test_CPU_Mat.rows-1,(int) matches.size()); k++) //THIS LOOP IS SENSITIVE TO SEGFAULTS
{
if((matches[k][0].distance < 0.6*(matches[k][1].distance)) && ((int) matches[k].size()<=2 && (int) matches[k].size()>0))
{
good_matches.push_back(matches[k][0]);
}
}
vector<Point2f> obj;
vector<Point2f> scene;
vector<Point2f> scene_corners(4);
Mat H;
Mat img (baseImages[i]), img_matches, frame_cpu (frame);
std::ostringstream o_stream;
o_stream<<"Logo_save/"<<baseImagesNames[i];
try
{
drawMatches( img, keypoints_tmp_CPU, frame_cpu, keypoints_test_CPU, good_matches, img_matches, Scalar::all(-1), Scalar::all(-1), vector<char>(), DrawMatchesFlags::NOT_DRAW_SINGLE_POINTS );
imwrite(o_stream.str(),img_matches);
}
catch(...)
{
cout<<"Error in drawMatches name: "<< baseImagesNames[i]<<endl;
}
if (good_matches.size() >= 4)
{
for( int k = 0; k < good_matches.size(); k++ )
{
//Get the keypoints from the good matches
obj.push_back( (keypoints_tmp_CPU)[ good_matches[k].queryIdx ].pt );
scene.push_back( keypoints_test_CPU[ good_matches[k].trainIdx ].pt );
}
cout<<good_matches.size()<<" "<<baseImagesNames[i]<<endl;
H = findHomography( obj, scene, CV_RANSAC);
perspectiveTransform( objs_corners, scene_corners, H);
bool falseDetect = isSmallAngle(scene_corners);
//cout<< falseDetect<< endl;
if(!falseDetect)
{
cout<<"DETECT "<<baseImagesNames[i]<<endl;
}
}
matcher.clear();
}
Bad result on GPU (MIN_HESSIAN==1500):
Bad result on GPU (MIN_HESSIAN==400):
CPU part:
SurfFeatureDetector detector( MIN_HESSIAN );//MIN_HESSIAN==400
const int baseImagesSize = baseImages.size();
vector< vector<KeyPoint> > kp_objects(baseImagesSize);
//Calculate descriptors (feature vectors)
SurfDescriptorExtractor extractor;
vector<Mat> des_objects(baseImagesSize);
FlannBasedMatcher matcher;
//namedWindow("SURF feature detector");
vector< vector<Point2f> > objs_corners(baseImagesSize,vector<Point2f>(4));
for (int i = 0; i < baseImagesSize; ++i)
{
detector.detect(baseImages[i], kp_objects[i]);
extractor.compute(baseImages[i], kp_objects[i], des_objects[i]);
//Get the corners from the object
(objs_corners[i])[0] = cvPoint(0,0);
(objs_corners[i])[1] = cvPoint( baseImages[i].cols, 0 );
(objs_corners[i])[2] = cvPoint( baseImages[i].cols, baseImages[i].rows );
(objs_corners[i])[3] = cvPoint( 0, baseImages[i].rows );
}
Mat des_image;
std::vector<KeyPoint> kp_image;
Mat image;
cvtColor(frame, image, CV_RGB2GRAY);
detector.detect( image, kp_image );
extractor.compute( image, kp_image, des_image );
for (int i = 0; i < baseImagesSize; ++i)
{
Mat img_matches;
std::vector<vector<DMatch > > matches;
std::vector<DMatch > good_matches;
std::vector<Point2f> obj;
std::vector<Point2f> scene;
std::vector<Point2f> scene_corners(4);
Mat H;
matcher.knnMatch(des_objects[i], des_image, matches, 2);
for(int k = 0; k < min(des_image.rows-1,(int) matches.size()); k++) //THIS LOOP IS SENSITIVE TO SEGFAULTS
{
if((matches[k][0].distance < 0.6*(matches[k][4].distance)) && ((int) matches[k].size()<=2 && (int) matches[k].size()>0))
{
good_matches.push_back(matches[k][0]);
}
}
//Draw only "good" matches
std::ostringstream o_stream;
o_stream<<"Logo_save/"<<baseImagesNames[i];
try
{
drawMatches( baseImages[i], kp_objects[i], image, kp_image, good_matches, img_matches, Scalar::all(-1), Scalar::all(-1), vector<char>(), DrawMatchesFlags::NOT_DRAW_SINGLE_POINTS );
imwrite(o_stream.str(),img_matches);
}
catch(...)
{
cout<<"Error in drawMatches name: "<< baseImagesNames[i]<<endl;
}
if (good_matches.size() >= 4)
{
for( int k = 0; k < good_matches.size(); k++ )
{
//Get the keypoints from the good matches
obj.push_back( (kp_objects[i])[ good_matches[k].queryIdx ].pt );
scene.push_back( kp_image[ good_matches[k].trainIdx ].pt );
}
H = findHomography( obj, scene, CV_RANSAC);
perspectiveTransform( objs_corners[i], scene_corners, H);
bool falseDetect = isSmallAngle(scene_corners);
if(!falseDetect)
{
cout<<"DETECT "<<baseImagesNames[i]<<endl;
}
}
}
Good result on CPU (MIN_HESSIAN==400):
Problem solved here topic
Swap parameters in drawMatches call:
drawMatches(frame_cpu, keypoints_test_CPU, img, keypoints_tmp_CPU, ...);
Swap keypoints_tmp_CPU and keypoints_test_CPU:
obj.push_back( keypoints_test_CPU[ good_matches[k].trainIdx ].pt );
scene.push_back( (keypoints_tmp_CPU)[ good_matches[k].queryIdx ].pt );