OpenCV - Image Stitching - image-processing

I am using following code to stitch to input images. For an unknown
reason the output result is crap!
It seems that the homography matrix is wrong (or is affected wrongly)
because the transformed image is like an "exploited star"!
I have commented the part that I guess is the source of the problem
but I cannot realize it.
Any help or point is appriciated!
Have a nice day,
Ali
void Stitch2Image(IplImage *mImage1, IplImage *mImage2)
{
// Convert input images to gray
IplImage* gray1 = cvCreateImage(cvSize(mImage1->width, mImage1->height), 8, 1);
cvCvtColor(mImage1, gray1, CV_BGR2GRAY);
IplImage* gray2 = cvCreateImage(cvSize(mImage2->width, mImage2->height), 8, 1);
cvCvtColor(mImage2, gray2, CV_BGR2GRAY);
// Convert gray images to Mat
Mat img1(gray1);
Mat img2(gray2);
// Detect FAST keypoints and BRIEF features in the first image
FastFeatureDetector detector(50);
BriefDescriptorExtractor descriptorExtractor;
BruteForceMatcher<L1<uchar> > descriptorMatcher;
vector<KeyPoint> keypoints1;
detector.detect( img1, keypoints1 );
Mat descriptors1;
descriptorExtractor.compute( img1, keypoints1, descriptors1 );
/* Detect FAST keypoints and BRIEF features in the second image*/
vector<KeyPoint> keypoints2;
detector.detect( img1, keypoints2 );
Mat descriptors2;
descriptorExtractor.compute( img2, keypoints2, descriptors2 );
vector<DMatch> matches;
descriptorMatcher.match(descriptors1, descriptors2, matches);
if (matches.size()==0)
return;
vector<Point2f> points1, points2;
for(size_t q = 0; q < matches.size(); q++)
{
points1.push_back(keypoints1[matches[q].queryIdx].pt);
points2.push_back(keypoints2[matches[q].trainIdx].pt);
}
// Create the result image
result = cvCreateImage(cvSize(mImage2->width * 2, mImage2->height), 8, 3);
cvZero(result);
// Copy the second image in the result image
cvSetImageROI(result, cvRect(mImage2->width, 0, mImage2->width, mImage2->height));
cvCopy(mImage2, result);
cvResetImageROI(result);
// Create warp image
IplImage* warpImage = cvCloneImage(result);
cvZero(warpImage);
/************************** Is there anything wrong here!? *******************/
// Find homography matrix
Mat H = findHomography(Mat(points1), Mat(points2), 8, 3.0);
CvMat HH = H; // Is this line converted correctly?
// Transform warp image
cvWarpPerspective(mImage1, warpImage, &HH);
// Blend
blend(result, warpImage);
/*******************************************************************************/
cvReleaseImage(&gray1);
cvReleaseImage(&gray2);
cvReleaseImage(&warpImage);
}

This is what I would suggest you to try, in this order:
1) Use CV_RANSAC option for homography. Refer http://opencv.willowgarage.com/documentation/cpp/calib3d_camera_calibration_and_3d_reconstruction.html
2) Try other descriptors, particularly SIFT or SURF which ship with OpenCV. For some images FAST or BRIEF descriptors are not discriminating enough. EDIT (Aug '12): The ORB descriptors, which are based on BRIEF, are quite good and fast!
3) Try to look at the Homography matrix (step through in debug mode or print it) and see if it is consistent.
4) If above does not give you a clue, try to look at the matches that are formed. Is it matching one point in one image with a number of points in the other image? If so the problem again should be with the descriptors or the detector.
My hunch is that it is the descriptors (so 1) or 2) should fix it).

Also switch to Hamming distance instead of L1 distance in BruteForceMatcher. BRIEF descriptors are supposed to be compared using Hamming distance.

Your homography, might calculated based on wrong matches and thus represent bad allignment.
I suggest to path the matrix through additional check of interdependancy between its rows.
You can use the following code:
bool cvExtCheckTransformValid(const Mat& T){
// Check the shape of the matrix
if (T.empty())
return false;
if (T.rows != 3)
return false;
if (T.cols != 3)
return false;
// Check for linear dependency.
Mat tmp;
T.row(0).copyTo(tmp);
tmp /= T.row(1);
Scalar mean;
Scalar stddev;
meanStdDev(tmp,mean,stddev);
double X = abs(stddev[0]/mean[0]);
printf("std of H:%g\n",X);
if (X < 0.8)
return false;
return true;
}

Related

OpenCV camera calibration with chessboard of different colours

A doubt came to my mind this morning: does the findChessboardCorners OpenCV function work with a chessboard of different colours, for example blue?
If it's not the case, do you think that a quite straightforward thresholding would do the trick?
You can't pass coloured images to the findChessboardCorners because it only takes a greyscale image as #api55 pointed out in his comment.
You might be worth taking a look at the checkchessboard code provided here
// does a fast check if a chessboard is in the input image. This is a workaround to
// a problem of cvFindChessboardCorners being slow on images with no chessboard
// - src: input binary image
// - size: chessboard size
// Returns 1 if a chessboard can be in this image and findChessboardCorners should be called,
// 0 if there is no chessboard, -1 in case of error
int checkChessboardBinary(const cv::Mat & img, const cv::Size & size)
{
CV_Assert(img.channels() == 1 && img.depth() == CV_8U);
Mat white = img.clone();
Mat black = img.clone();
int result = 0;
for ( int erosion_count = 0; erosion_count <= 3; erosion_count++ )
{
if ( 1 == result )
break;
if ( 0 != erosion_count ) // first iteration keeps original images
{
erode(white, white, Mat(), Point(-1, -1), 1);
dilate(black, black, Mat(), Point(-1, -1), 1);
}
vector<pair<float, int> > quads;
fillQuads(white, black, 128, 128, quads);
if (checkQuads(quads, size))
result = 1;
}
return result;
}
With the main loop being:
CV_IMPL
int cvFindChessboardCorners( const void* arr, CvSize pattern_size,
CvPoint2D32f* out_corners, int* out_corner_count,
int flags )
is the main implementation of this method. In here they
Use cvCheckChessboard to determine if a chessboard is in the image
Convert to binary (B&W) and dilate to split the corners apart Use
icvGenerateQuads to find the squares.
So in answer to your question, as long as there is sufficient contrast in your image after you convert it to greyscale it will likely work, I would imagine a greyscaled blue and white image would be good enough, if it was a light aqua or yellow or something you might struggle without more processing

Is it possible to recognize so minimal changes between noisy images in OpenCV?

I want to detect the very minimal movement of a conveyor belt using image evaluation (Resolution: 31x512, image rate: 1000 per second.). The moment of belt-start is important for me.
If I do cv::absdiff between two subsequent images, I obtain very noisy result:
According to the mechanical rotation sensor of the motor, the movement starts here:
I tried to threshold the abs-diff image with a cascade of erosion and dilation, but I could detect the earliest change more than second too late in this image:
Is it possible to find the change earlier?
Here is the sequence of the Images without changes (according to motor sensor):
In this sequence the movement begins in the middle image:
Looks like I've found a solution which works in MY case.
Instead of comparing the image changes in space-domain, the cross-correlation should be applied:
I convert both images to DFT, multiply DFT-Mats and convert back. The max pixel value is the center of the correlation. As long as the images are same, the max-pix remains in the same position and moves otherwise.
The actual working code uses 3 images, 2 DFT multiplication result between images 1,2 and 2,3:
Mat img1_( 512, 32, CV_16UC1 );
Mat img2_( 512, 32, CV_16UC1 );
Mat img3_( 512, 32, CV_16UC1 );
//read the data in the images wohever you want. I read from MHD-file
//Set ROI (if required)
Mat img1 = img1_(cv::Rect(0,200,32,100));
Mat img2 = img2_(cv::Rect(0,200,32,100));
Mat img3 = img3_(cv::Rect(0,200,32,100));
//Float mats for DFT
Mat img1f;
Mat img2f;
Mat img3f;
//DFT and produtcts mats
Mat dft1,dft2,dft3,dftproduct,dftproduct2;
//Calculate DFT of both images
img1.convertTo(img1f, CV_32FC1);
cv::dft(img1f, dft1);
img2.convertTo(img3f, CV_32FC1);
cv::dft(img3f, dft3);
img3.convertTo(img2f, CV_32FC1);
cv::dft(img2f, dft2);
//Multiply DFT Mats
cv::mulSpectrums(dft1,dft2,dftproduct,true);
cv::mulSpectrums(dft2,dft3,dftproduct2,true);
//Convert back to space domain
cv::Mat result,result2;
cv::idft(dftproduct,result);
cv::idft(dftproduct2,result2);
//Not sure if required, I needed it for visualizing
cv::normalize( result, result, 0, 255, NORM_MINMAX, CV_8UC1);
cv::normalize( result2, result2, 0, 255, NORM_MINMAX, CV_8UC1);
//Find maxima positions
double dummy;
Point locdummy; Point maxLoc1; Point maxLoc2;
cv::minMaxLoc(result, &dummy, &dummy, &locdummy, &maxLoc1);
cv::minMaxLoc(result2, &dummy, &dummy, &locdummy, &maxLoc2);
//Calculate products simply fot having one value to compare
int maxlocProd1 = maxLoc1.x*maxLoc1.y;
int maxlocProd2 = maxLoc2.x*maxLoc2.y;
//Calculate absolute difference of the products. Not 0 means movement
int absPosDiff = std::abs(maxlocProd2-maxlocProd1);
if ( absPosDiff>0 )
{
std::cout << id<< std::endl;
break;
}

Accuracy tuning for Haar-Cascade Classifier

I'm using Haar-Cascade Classifier in order to detect faces.
I'm currently facing some problems with the following function:
void ImageManager::detectAndDisplay(Mat frame, CascadeClassifier face_cascade){
string window_name = "Capture - Face detection";
string filename;
std::vector<Rect> faces;
std::vector<Rect> eyes;
Mat frame_gray;
Mat crop;
Mat res;
Mat gray;
string text;
stringstream sstm;
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));
// Set Region of Interest
cv::Rect roi_b;
cv::Rect roi_c;
size_t ic = 0; // ic is index of current element
for (ic = 0; ic < faces.size(); ic++) // Iterate through all current elements (detected faces)
{
roi_c.x = faces[ic].x;
roi_c.y = faces[ic].y;
roi_c.width = (faces[ic].width);
roi_c.height = (faces[ic].height);
crop = frame_gray(roi_c);
faces_img.push_back(crop);
rectangle(frame, Point(roi_c.x, roi_c.y), Point(roi_c.x + roi_c.width, roi_c.y + roi_c.height), Scalar(0,0,255), 2);
}
imshow("test", frame);
waitKey(0);
cout << faces_img.size();
}
The frame is the photo I'm trying to scan.
The face_cascade is the classifier.
internally, the CascadeClassifier does several detections, and groups those.
minNeighbours (in the detectMultiScale call) is the amount of detections in about the same place nessecary to count as a valid detection, so increase that from your current 2 to maybe 5 or so, until you start to miss positives.
As an addition to berak's statement, it's not only about reducing/increasing of detectMultiScale parameters if you're not doing the stuff only on an image. You'll face performance problems that do not let the user use the application.
Performance issues are relying on miscalculations. And what calculation takes is just testing.
If you are not trying to have the best results under different light conditions(since this is visual-dependent information) you'll have to scale the input array before sending it as an argument to detectMultiScale function. Once detection's completed, rescale to the previous size(it may be done by changing the rectangle's size that's used as an argument for detectMultiScale).

Having some difficulty in image stitching using OpenCV

I'm currently working on Image stitching using OpenCV 2.3.1 on Visual Studio 2010, but I'm having some trouble.
Problem Description
I'm trying to write a code for stitching multiple images derived from a few cameras(about 3~4), i,e, the code should keep executing image stitching until I ask it to stop.
The following is what I've done so far:
(For simplification, I'll replace some part of the code with just a few words)
1.Reading frames(images) from 2 cameras (Currently I'm just working on 2 cameras.)
2.Feature detection, descriptor calculation (SURF)
3.Feature matching using FlannBasedMatcher
4.Removing outliers and calculate the Homography with inliers using RANSAC.
5.Warp one of both images.
For step 5., I followed the answer in the following thread and just changed some parameters:
Stitching 2 images in opencv
However, the result is terrible though.
I just uploaded the result onto youtube and of course only those who have the link will be able to see it.
http://youtu.be/Oy5z_7LeaMk
My code is shown below:
(Only crucial parts are shown)
VideoCapture cam1, cam2;
cam1.open(0);
cam2.open(1);
while(1)
{
Mat frm1, frm2;
cam1 >> frm1;
cam2 >> frm2;
//(SURF detection, descriptor calculation
//and matching using FlannBasedMatcher)
double max_dist = 0; double min_dist = 100;
//-- Quick calculation of max and min distances between keypoints
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;
}
(Draw only "good" matches
(i.e. whose distance is less than 3*min_dist ))
vector<Point2f> frame1;
vector<Point2f> frame2;
for( int i = 0; i < good_matches.size(); i++ )
{
//-- Get the keypoints from the good matches
frame1.push_back( keypoints_1[ good_matches[i].queryIdx ].pt );
frame2.push_back( keypoints_2[ good_matches[i].trainIdx ].pt );
}
Mat H = findHomography( Mat(frame1), Mat(frame2), CV_RANSAC );
cout << "Homography: " << H << endl;
/* warp the image */
Mat warpImage2;
warpPerspective(frm2, warpImage2,
H, Size(frm2.cols, frm2.rows), INTER_CUBIC);
Mat final(Size(frm2.cols*3 + frm1.cols, frm2.rows),CV_8UC3);
Mat roi1(final, Rect(frm1.cols, 0, frm1.cols, frm1.rows));
Mat roi2(final, Rect(2*frm1.cols, 0, frm2.cols, frm2.rows));
warpImage2.copyTo(roi2);
frm1.copyTo(roi1);
imshow("final", final);
What else should I do to make the stitching better?
Besides, is it reasonable to make the Homography matrix fixed instead of keeping computing it ?
What I mean is to specify the angle and the displacement between the 2 cameras by myself so as to derive a Homography matrix that satisfies what I want.
Thanks. :)
It sounds like you are going about this sensibly, but if you have access to both of the cameras, and they will remain stationary with respect to each other, then calibrating offline, and simply applying the transformation online will make your application more efficient.
One point to note is, you say you are using the findHomography function from OpenCV. From the documentation, this function:
Finds a perspective transformation between two planes.
However, your points are not restricted to a specific plane as they are imaging a 3D scene. If you wanted to calibrate offline, you could image a chessboard with both cameras, and the detected corners could be used in this function.
Alternatively, you may like to investigate the Fundamental matrix, which can be calculated with a similar function. This matrix describes the relative position of the cameras, but some work (and a good textbook) will be required to extract them.
If you can find it, I would strongly recommend having a look at Part II: "Two-View Geometry" in the book "Multiple View Geometry in computer vision", by Richard Hartley and Andrew Zisserman, which goes through the process in detail.
I have been working lately on image registration. My algorithm takes two images, calculates the SURF features, find correspondences, find homography matrix and then stitch both images together, I did it with the next code:
void stich(Mat base, Mat target,Mat homography, Mat& panorama){
Mat corners1(1, 4,CV_32F);
Mat corners2(1,4,CV_32F);
Mat corners(1,4,CV_32F);
vector<Mat> planes;
/* compute corners
of warped image
*/
corners1.at<float>(0,0)=0;
corners2.at<float>(0,0)=0;
corners1.at<float>(0,1)=0;
corners2.at<float>(0,1)=target.rows;
corners1.at<float>(0,2)=target.cols;
corners2.at<float>(0,2)=0;
corners1.at<float>(0,3)=target.cols;
corners2.at<float>(0,3)=target.rows;
planes.push_back(corners1);
planes.push_back(corners2);
merge(planes,corners);
perspectiveTransform(corners, corners, homography);
/* compute size of resulting
image and allocate memory
*/
double x_start = min( min( (double)corners.at<Vec2f>(0,0)[0], (double)corners.at<Vec2f> (0,1)[0]),0.0);
double x_end = max( max( (double)corners.at<Vec2f>(0,2)[0], (double)corners.at<Vec2f>(0,3)[0]), (double)base.cols);
double y_start = min( min( (double)corners.at<Vec2f>(0,0)[1], (double)corners.at<Vec2f>(0,2)[1]), 0.0);
double y_end = max( max( (double)corners.at<Vec2f>(0,1)[1], (double)corners.at<Vec2f>(0,3)[1]), (double)base.rows);
/*Creating image
with same channels, depth
as target
and proper size
*/
panorama.create(Size(x_end - x_start + 1, y_end - y_start + 1), target.depth());
planes.clear();
/*Planes should
have same n.channels
as target
*/
for (int i=0;i<target.channels();i++){
planes.push_back(panorama);
}
merge(planes,panorama);
// create translation matrix in order to copy both images to correct places
Mat T;
T=Mat::zeros(3,3,CV_64F);
T.at<double>(0,0)=1;
T.at<double>(1,1)=1;
T.at<double>(2,2)=1;
T.at<double>(0,2)=-x_start;
T.at<double>(1,2)=-y_start;
// copy base image to correct position within output image
warpPerspective(base, panorama, T,panorama.size(),INTER_LINEAR| CV_WARP_FILL_OUTLIERS);
// change homography to take necessary translation into account
gemm(T, homography,1,T,0,T);
// warp second image and copy it to output image
warpPerspective(target,panorama, T, panorama.size(),INTER_LINEAR);
//tidy
corners.release();
T.release();
}
Any question I will try

OpenCV Surf and Outliers detection

I know there are already several questions with the same subject asked here, but I couldn't find any help.
So I want to compare 2 images in order to see how similar they are and I'm using the well known find_obj.cpp demo to extract surf descriptors and then for the matching I use the flannFindPairs.
But as you know this method doesn't discard the outliers and I'd like to know the number of true positive matches so I can figure how similar those two images are.
I have already seen this question: Detecting outliers in SURF or SIFT algorithm with OpenCV and the guy there suggests to use the findFundamentalMat but once you get the fundamental matrix how can I get the number of outliers/true positive from that matrix? Thank you.
Here is a snippet from the descriptor_extractor_matcher.cpp sample available from OpenCV:
if( !isWarpPerspective && ransacReprojThreshold >= 0 )
{
cout << "< Computing homography (RANSAC)..." << endl;
vector<Point2f> points1; KeyPoint::convert(keypoints1, points1, queryIdxs);
vector<Point2f> points2; KeyPoint::convert(keypoints2, points2, trainIdxs);
H12 = findHomography( Mat(points1), Mat(points2), CV_RANSAC, ransacReprojThreshold );
cout << ">" << endl;
}
Mat drawImg;
if( !H12.empty() ) // filter outliers
{
vector<char> matchesMask( filteredMatches.size(), 0 );
vector<Point2f> points1; KeyPoint::convert(keypoints1, points1, queryIdxs);
vector<Point2f> points2; KeyPoint::convert(keypoints2, points2, trainIdxs);
Mat points1t; perspectiveTransform(Mat(points1), points1t, H12);
double maxInlierDist = ransacReprojThreshold < 0 ? 3 : ransacReprojThreshold;
for( size_t i1 = 0; i1 < points1.size(); i1++ )
{
if( norm(points2[i1] - points1t.at<Point2f>((int)i1,0)) <= maxInlierDist ) // inlier
matchesMask[i1] = 1;
}
// draw inliers
drawMatches( img1, keypoints1, img2, keypoints2, filteredMatches, drawImg, CV_RGB(0, 255, 0), CV_RGB(0, 0, 255), matchesMask
#if DRAW_RICH_KEYPOINTS_MODE
, DrawMatchesFlags::DRAW_RICH_KEYPOINTS
#endif
);
#if DRAW_OUTLIERS_MODE
// draw outliers
for( size_t i1 = 0; i1 < matchesMask.size(); i1++ )
matchesMask[i1] = !matchesMask[i1];
drawMatches( img1, keypoints1, img2, keypoints2, filteredMatches, drawImg, CV_RGB(0, 0, 255), CV_RGB(255, 0, 0), matchesMask,
DrawMatchesFlags::DRAW_OVER_OUTIMG | DrawMatchesFlags::NOT_DRAW_SINGLE_POINTS );
#endif
}
else
drawMatches( img1, keypoints1, img2, keypoints2, filteredMatches, drawImg );
The key lines for the filtering are performed here:
if( norm(points2[i1] - points1t.at<Point2f>((int)i1,0)) <= maxInlierDist ) // inlier
matchesMask[i1] = 1;
Which is measuring the L2-norm distance between the points (either 3 pixels if nothing was specified, or user-defined number of pixels reprojection error).
Hope that helps!
you can use the size of the vector named "ptpairs" in order to decide how similiar the pictures are.
this vector contains the matching keypoints, so his size/2 is the number of matches.
i think you can use the size of ptpairs divided by the total number of keypoints in order to set an appropriate threshold.
this will probably give you an estimation to the similiarty between them.

Resources