I am working with a fish-eye camera and need the reverse the distortion before any further calculation,
In this question this is happening Correcting fisheye distortion
src = cv.LoadImage(src)
dst = cv.CreateImage(cv.GetSize(src), src.depth, src.nChannels)
mapx = cv.CreateImage(cv.GetSize(src), cv.IPL_DEPTH_32F, 1)
mapy = cv.CreateImage(cv.GetSize(src), cv.IPL_DEPTH_32F, 1)
cv.InitUndistortMap(intrinsics, dist_coeffs, mapx, mapy)
cv.Remap(src, dst, mapx, mapy, cv.CV_INTER_LINEAR + cv.CV_WARP_FILL_OUTLIERS, cv.ScalarAll(0))
The problem with this is that this way the remap functions goes through all the points and creates a new picture out of. this is time consuming to do it every frame.
They way that I am looking for is to have a point to point translation on the fish-eye picture to normal picture coordinates.
The approach we are taking is to do all the calculations on the input frame and just translate the result coordinates to the world coordinates so we don't want to go through all the points of a picture and create a new one out of it. (Time is really important for us)
In the matrices mapx and mapy there are some point to point translations but a lot of points are without complete translation.
I tried to interpolate this matrices but the result was not what I was looking for.
Any help in would be much appreciated, even other approaches which are more time efficient than cv.Remap.
Thanks
I think what you want is cv.UndistortPoints().
Assuming you have detected some point features distorted in your distorted image, you should be able to do something like this:
cv.UndistortPoints(distorted, undistorted, intrinsics, dist_coeffs)
This will allow you to work with undistorted points without generating a new, undistorted image for each frame.
Related
I'm trying to deskew an image that has an element of known size. Given this image:
I can use aruco:: estimatePoseBoard which returns rotation and translation vectors. Is there a way to use that information to deskew everything that's in the same plane as the marker board? (Unfortunately my linear algebra is rudimentary at best.)
Clarification
I know how to deskew the marker board. What I want to be able to do is deskew the other things (in this case, the cloud-shaped object) in the same plane as the marker board. I'm trying to determine whether or not that's possible and, if so, how to do it. I can already put four markers around the object I want to deskew and use the detected corners as input to getPerspectiveTransform along with the known distance between them. But for our real-world application it may be difficult for the user to place markers exactly. It would be much easier if they could place a single marker board in the frame and have the software deskew the other objects.
Since you tagged OpenCV:
From the image I can see that you have detected the corners of all the black box. So just get the most border for points in a way or another:
Then it is like this:
std::vector<cv::Point2f> src_points={/*Fill your 4 corners here*/};
std::vector<cv::Point2f> dst_points={cv:Point2f(0,0), cv::Point2f(width,0), cv::Point2f(width,height),cv::Point2f(0,height)};
auto H=v::getPerspectiveTransform(src_points,dst_points);
cv::Mat copped_image;
cv::warpPerspective(full_image,copped_image,H,cv::Size(width,height));
I was stuck on the assumption that the destination points in the call to getPerspectiveTransform had to be the corners of the output image (as they are in Humam's suggestion). Once it dawned on me that the destination points could be somewhere within the output image I had my answer.
float boardX = 1240;
float boardY = 1570;
float boardWidth = 1730;
float boardHeight = 1400;
vector<Point2f> destinationCorners;
destinationCorners(Point2f(boardX+boardWidth, boardY));
destinationCorners(Point2f(boardX+boardWidth, boardY+boardHeight));
destinationCorners(Point2f(boardX, boardY+boardHeight));
destinationCorners(Point2f(boardX, boardY));
Mat h = getPerspectiveTransform(detectedCorners, destinationCorners);
Mat bigImage(image.size() * 3, image.type(), Scalar(0, 50, 50));
warpPerspective(image, bigImage, h, bigImage.size());
This fixed the perspective of the board and everything in its plane. (The waviness of the board is due to the fact that the paper wasn't lying flat in the original photo.)
I'm getting results I don't expect when I use OpenCV 3.0 calibrateCamera. Here is my algorithm:
Load in 30 image points
Load in 30 corresponding world points (coplanar in this case)
Use points to calibrate the camera, just for un-distorting
Un-distort the image points, but don't use the intrinsics (coplanar world points, so intrinsics are dodgy)
Use the undistorted points to find a homography, transforming to world points (can do this because they are all coplanar)
Use the homography and perspective transform to map the undistorted points to the world space
Compare the original world points to the mapped points
The points I have are noisy and only a small section of the image. There are 30 coplanar points from a single view so I can't get camera intrinsics, but should be able to get distortion coefficients and a homography to create a fronto-parallel view.
As expected, the error varies depending on the calibration flags. However, it varies opposite to what I expected. If I allow all variables to adjust, I would expect error to come down. I am not saying I expect a better model; I actually expect over-fitting, but that should still reduce error. What I see though is that the fewer variables I use, the lower my error. The best result is with a straight homography.
I have two suspected causes, but they seem unlikely and I'd like to hear an unadulterated answer before I air them. I have pulled out the code to just do what I'm talking about. It's a bit long, but it includes loading the points.
The code doesn't appear to have bugs; I've used "better" points and it works perfectly. I want to emphasize that the solution here can't be to use better points or perform a better calibration; the whole point of the exercise is to see how the various calibration models respond to different qualities of calibration data.
Any ideas?
Added
To be clear, I know the results will be bad and I expect that. I also understand that I may learn bad distortion parameters which leads to worse results when testing points that have not been used to train the model. What I don't understand is how the distortion model has more error when using the training set as the test set. That is, if the cv::calibrateCamera is supposed to choose parameters to reduce error over the training set of points provided, yet it is producing more error than if it had just selected 0s for K!, K2, ... K6, P1, P2. Bad data or not, it should at least do better on the training set. Before I can say the data is not appropriate for this model, I have to be sure I'm doing the best I can with the data available, and I can't say that at this stage.
Here an example image
The points with the green pins are marked. This is obviously just a test image.
Here is more example stuff
In the following the image is cropped from the big one above. The centre has not changed. This is what happens when I undistort with just the points marked manually from the green pins and allowing K1 (only K1) to vary from 0:
Before
After
I would put it down to a bug, but when I use a larger set of points that covers more of the screen, even from a single plane, it works reasonably well. This looks terrible. However, the error is not nearly as bad as you might think from looking at the picture.
// Load image points
std::vector<cv::Point2f> im_points;
im_points.push_back(cv::Point2f(1206, 1454));
im_points.push_back(cv::Point2f(1245, 1443));
im_points.push_back(cv::Point2f(1284, 1429));
im_points.push_back(cv::Point2f(1315, 1456));
im_points.push_back(cv::Point2f(1352, 1443));
im_points.push_back(cv::Point2f(1383, 1431));
im_points.push_back(cv::Point2f(1431, 1458));
im_points.push_back(cv::Point2f(1463, 1445));
im_points.push_back(cv::Point2f(1489, 1432));
im_points.push_back(cv::Point2f(1550, 1461));
im_points.push_back(cv::Point2f(1574, 1447));
im_points.push_back(cv::Point2f(1597, 1434));
im_points.push_back(cv::Point2f(1673, 1463));
im_points.push_back(cv::Point2f(1691, 1449));
im_points.push_back(cv::Point2f(1708, 1436));
im_points.push_back(cv::Point2f(1798, 1464));
im_points.push_back(cv::Point2f(1809, 1451));
im_points.push_back(cv::Point2f(1819, 1438));
im_points.push_back(cv::Point2f(1925, 1467));
im_points.push_back(cv::Point2f(1929, 1454));
im_points.push_back(cv::Point2f(1935, 1440));
im_points.push_back(cv::Point2f(2054, 1470));
im_points.push_back(cv::Point2f(2052, 1456));
im_points.push_back(cv::Point2f(2051, 1443));
im_points.push_back(cv::Point2f(2182, 1474));
im_points.push_back(cv::Point2f(2171, 1459));
im_points.push_back(cv::Point2f(2164, 1446));
im_points.push_back(cv::Point2f(2306, 1474));
im_points.push_back(cv::Point2f(2292, 1462));
im_points.push_back(cv::Point2f(2278, 1449));
// Create corresponding world / object points
std::vector<cv::Point3f> world_points;
for (int i = 0; i < 30; i++) {
world_points.push_back(cv::Point3f(5 * (i / 3), 4 * (i % 3), 0.0f));
}
// Perform calibration
// Flags are set out so they can be commented out and "freed" easily
int calibration_flags = 0
| cv::CALIB_FIX_K1
| cv::CALIB_FIX_K2
| cv::CALIB_FIX_K3
| cv::CALIB_FIX_K4
| cv::CALIB_FIX_K5
| cv::CALIB_FIX_K6
| cv::CALIB_ZERO_TANGENT_DIST
| 0;
// Initialise matrix
cv::Mat intrinsic_matrix = cv::Mat(3, 3, CV_64F);
intrinsic_matrix.ptr<float>(0)[0] = 1;
intrinsic_matrix.ptr<float>(1)[1] = 1;
cv::Mat distortion_coeffs = cv::Mat::zeros(5, 1, CV_64F);
// Rotation and translation vectors
std::vector<cv::Mat> undistort_rvecs;
std::vector<cv::Mat> undistort_tvecs;
// Wrap in an outer vector for calibration
std::vector<std::vector<cv::Point2f>>im_points_v(1, im_points);
std::vector<std::vector<cv::Point3f>>w_points_v(1, world_points);
// Calibrate; only 1 plane, so intrinsics can't be trusted
cv::Size image_size(4000, 3000);
calibrateCamera(w_points_v, im_points_v,
image_size, intrinsic_matrix, distortion_coeffs,
undistort_rvecs, undistort_tvecs, calibration_flags);
// Undistort im_points
std::vector<cv::Point2f> ud_points;
cv::undistortPoints(im_points, ud_points, intrinsic_matrix, distortion_coeffs);
// ud_points have been "unintrinsiced", but we don't know the intrinsics, so reverse that
double fx = intrinsic_matrix.at<double>(0, 0);
double fy = intrinsic_matrix.at<double>(1, 1);
double cx = intrinsic_matrix.at<double>(0, 2);
double cy = intrinsic_matrix.at<double>(1, 2);
for (std::vector<cv::Point2f>::iterator iter = ud_points.begin(); iter != ud_points.end(); iter++) {
iter->x = iter->x * fx + cx;
iter->y = iter->y * fy + cy;
}
// Find a homography mapping the undistorted points to the known world points, ground plane
cv::Mat homography = cv::findHomography(ud_points, world_points);
// Transform the undistorted image points to the world points (2d only, but z is constant)
std::vector<cv::Point2f> estimated_world_points;
std::cout << "homography" << homography << std::endl;
cv::perspectiveTransform(ud_points, estimated_world_points, homography);
// Work out error
double sum_sq_error = 0;
for (int i = 0; i < 30; i++) {
double err_x = estimated_world_points.at(i).x - world_points.at(i).x;
double err_y = estimated_world_points.at(i).y - world_points.at(i).y;
sum_sq_error += err_x*err_x + err_y*err_y;
}
std::cout << "Sum squared error is: " << sum_sq_error << std::endl;
I would take random samples of the 30 input points and compute the homography in each case along with the errors under the estimated homographies, a RANSAC scheme, and verify consensus between error levels and homography parameters, this can be just a verification of the global optimisation process. I know that might seem unnecessary, but it is just a sanity check for how sensitive the procedure is to the input (noise levels, location)
Also, it seems logical that fixing most of the variables gets you the least errors, as the degrees of freedom in the minimization process are less. I would try fixing different ones to establish another consensus. At least this would let you know which variables are the most sensitive to the noise levels of the input.
Hopefully, such a small section of the image would be close to the image centre as it will incur the least amount of lens distortion. Is using a different distortion model possible in your case? A more viable way is to adapt the number of distortion parameters given the position of the pattern with respect to the image centre.
Without knowing the constraints of the algorithm, I might have misunderstood the question, that's also an option too, in such case I can roll back.
I would like to have this as a comment rather, but I do not have enough points.
OpenCV runs Levenberg-Marquardt algorithm inside calibrate camera.
https://en.wikipedia.org/wiki/Levenberg%E2%80%93Marquardt_algorithm/
This algortihm works fine in problems with one minimum. In case of single image, points located close each other and many dimensional problem (n= number of coefficents) algorithm may be unstable (especially with wrong initial guess of camera matrix. Convergence of algorithm is well described here:
https://na.math.kit.edu/download/papers/levenberg.pdf/
As you wrote, error depends on calibration flags - number of flags changes dimension of a problem to be optimized.
Camera calibration also calculates pose of camera, which will be bad in models with wrong calibration matrix.
As a solution I suggest changing approach. You dont need to calculate camera matrix and pose in this step. Since you know, that points are located on a plane you can use 3d-2d plane projection equation to determine distribution type of points. By distribution I mean, that all points will be located equally on some kind of trapezoid.
Then you can use cv::undistort with different distCoeffs on your test image and calculate image point distribution and distribution error.
The last step will be to perform this steps as a target function for some optimization algorithm with distortion coefficents being optimized.
This is not the easiest solution, but i hope it will help you.
I calibrated a wide angle camera (~120° horizontal, ~110° vertical) using the Pinhole camera model. The result is quite good but large areas of the image are discarded.
So I'm trying to use the getOptimalNewCameraMatrix function (using alpha = 1) to calculate the newCameraMatrix and then pass it to cv::undistord, but the resulting frame is really distorted.
The code I use is:
// Matrices calculation
data.rms = calibrateCamera(objectPoints, imagePoints, data.size, data.cameraMatrix,
data.distCoeffs, data.rvecs, data.tvecs, settings.calibFlags);
if(!settings.cropCorrected) {
Rect roi;
data.newCameraMatrix = getOptimalNewCameraMatrix(data.cameraMatrix, data.distCoeffs,
data.size, 1, data.size, &roi, true);
}
// Undistort
cv::undistort(in, corrected, cameraMatrix, distCoeffs, newCameraMatrix);
This is simple but there is something not working. I have already tried to use roi to crop the inner area, in the hope that it would have been anyway bigger than without the additional matrix, but in fact it's not.
How can i resolve the situation?
Ok, I figured it out. The problem is that alpha = 1.0 is simply too big. Curiously even using 0.0 the newCameraMatrix enlarges by ~20° the fields of view and apparently no wrong pixels or distortion are present.
I am currently lost in the OpenCV documentation and am looking for some guidance on the possible ordering of functions, or perhaps a function within OpenCV that I haven't came acrossed yet...
I am tracking a laser blob within a camera feed to a location on a projection screen. Up until now I have been using findHomography and then projectTransform to accomplish this however the camera I was using had very little distortion. Now I am using a different camera with a noticeable radial distortion. I have used cvCalibrateCamera to get the distortion coefficients, camera matrix, etc. but I am not sure how I should use this data with my current process, or perhaps I need to use different functions and/or ordering of functions from OpenCV altogether. Any suggestions would be appreciated...
My current code that works well (without distortion) is as follows:
Mat homog;
homog = findHomography(Mat(vCameraPoints), Mat(vTargetPoints), CV_RANSAC);
vector<Point2f> cvTrackPoint;
cvTrackPoint.push_back(Point2f(pMapPoint.fX, pMapPoint.fY));
Mat normalizedImageMat;
perspectiveTransform(Mat(cvTrackPoint), normalizedImageMat, homog);
Point2f normalizedImgPt;
normalizedImgPt = Point2f(normalizedImageMat.at<Point2f>(0,0));
normalizedImgPt.x /= szCameraSize.fWidth;
normalizedImgPt.y /= szCameraSize.fHeight;
I then of course multiply the normalizedImgPt to my projection screen resolution
So again, just to clarify...I do have what appears to be good data from calibrateCamera, how would I use this information to factor in the lens distortion? Perhaps the above process wont work, any help?
Thanks, in advance
If you have acquired the distortion coefficients, then a simple (yet probably suboptimal) way to get back to the non-distorted case would be to undistort the image. The undistorted image is the image a camera with similar intrinsic and extrinsic parameters but without lens distorsion would acquire.
The corresponding OpenCV function is undistort
I have only just started experimenting with OpenCV a little bit. I have a setup of an LCD with a static position, and I'd like to extract what is being displayed on the screen from the image. I've seen the chessboard pattern used for calibrating a camera, but it seems like that is used to undistort the image, which isn't totally what I want to do.
I was thinking I'd display the chessboard on the LCD and then figure out the transformations needed to convert the image of the LCD into the ideal view of the chessboard directly overhead and cropped. Then I would store the transformations, change what the LCD is displaying, take a picture, perform the same transformations, and get the ideal view of what was now being displayed.
I'm wondering if that sounds like a good idea? Is there a simpler way to achieve what I'm trying to do? And any tips on the functions I should be using to figure out the transformations, perform them, store them (maybe just keep the transform matrices in memory or write them to file), etc?
I'm not sure I understood correctly everything you are trying to do, but bear with me.
Some cameras have lenses that cause a little distortion to the image, and for this purpose OpenCV offers methods to aid in the camera calibration process.
Practically speaking, if you want to write an application that will automatically correct the distortion in the image, first, you need to discover what are the magical values that need to be used to undo this effect. These values come from a proper calibration procedure.
The chessboard image is used together with an application to calibrate the camera. So, after you have an image of the chessboard taken by the camera device, pass this image to the calibration app. The app will identify the corners of the squares and compute the values of the distortion and return the magical values you need to use to counter the distortion effect. At this point, you are interested in 2 variables returned by calibrateCamera(): they are cameraMatrix and distCoeffs. Print them, and write the data on a piece of paper.
At the end, the system you are developing needs to have a function/method to undistort the image, where these 2 variables will be hard coded inside the function, followed by a call to cv::undistort() (if you are using the C++ API of OpenCV):
cv::Mat undistorted;
cv::undistort(image, undistorted, cameraMatrix, distCoeffs);
and that's it.
Detecting rotation automatically might be a bit tricky, but the first thing to do is find the coordinates of the object you are interested in. But if the camera is in a fixed position, this is going to be easy.
For more info on perspective change and rotation with OpenCV, I suggest taking a look at these other questions:
Executing cv::warpPerspective for a fake deskewing on a set of cv::Point
Affine Transform, Simple Rotation and Scaling or something else entirely?
Rotate cv::Mat using cv::warpAffine offsets destination image
findhomography() is not bad choice, but skew,distortion(camera lens) is real problem..
C++: Mat findHomography(InputArray srcPoints, InputArray dstPoints,
int method=0, double ransacReprojThreshold=3, OutputArray
mask=noArray() )
Python: cv2.findHomography(srcPoints, dstPoints[, method[,
ransacReprojThreshold[, mask]]]) → retval, mask
C: void cvFindHomography(const CvMat* srcPoints, const CvMat*
dstPoints, CvMat* H, int method=0, double ransacReprojThreshold=3,
CvMat* status=NULL)
http://opencv.itseez.com/modules/calib3d/doc/camera_calibration_and_3d_reconstruction.html#findhomography