antithing's profile - activity

Opencv function crashes on AMD processors, runs fine on Intel. As above, I have a function in my application that uses o

Opencv function crashes on AMD processors, runs fine on Intel. As above, I have a function in my application that uses o

Opencv function crashes on AMD processors, runs fine on Intel. As above, I have a function in my application that uses o

Find point correspondences without features or descriptors? I have a set of data that is as follows: std::vector<cv:

Video IO backends, ffmpeg vs I am developing an application which will be sold commercially. It requires video file loa

Is this the correct approach? //test normailse points std::vector<cv::Point2f> outputUndistortedPoints; c

Hi, thank you for getting back to me. Can you please explain the second option for me? How do I compute the image coordi

Using findHomography to create 'fake' intrinsics? I have two images from very different cameras. cv::Mat CameraA is fr

Using findHomography to create 'fake' intrinsics? I have two images from very different cameras. cv::Mat CameraA is fro

I am using the following code, pulled straight from the tutorials:

void detectCharucoBoardWithoutCalibration(cv::Mat image)
{

    cv::Ptr<cv::aruco::Dictionary> dictionary = cv::aruco::getPredefinedDictionary(cv::aruco::DICT_6X6_1000);
    cv::Ptr<cv::aruco::CharucoBoard> board = cv::aruco::CharucoBoard::create(7, 5, 0.04f, 0.02f, dictionary);
    cv::Ptr<cv::aruco::DetectorParameters> params = cv::aruco::DetectorParameters::create();
    params->cornerRefinementMethod = cv::aruco::CORNER_REFINE_NONE;
        cv::Mat  imageCopy;
        image.copyTo(imageCopy);
        std::vector<int> markerIds;
        std::vector<std::vector<cv::Point2f> > markerCorners;
        cv::aruco::detectMarkers(image, board->dictionary, markerCorners, markerIds, params);
        //or
        //cv::aruco::detectMarkers(image, dictionary, markerCorners, markerIds, params);
        // if at least one marker detected
        if (markerIds.size() > 0) {
            cv::aruco::drawDetectedMarkers(imageCopy, markerCorners, markerIds);
            std::vector<cv::Point2f> charucoCorners;
            std::vector<int> charucoIds;
            cv::aruco::interpolateCornersCharuco(markerCorners, markerIds, image, board, charucoCorners, charucoIds);
            // if at least one charuco corner detected
            if (charucoIds.size() > 0)
                cv::aruco::drawDetectedCornersCharuco(imageCopy, charucoCorners, charucoIds, cv::Scalar(255, 0, 0));
        }
        cv::imshow("out", imageCopy);
        cv::imwrite("data/markers.jpg", imageCopy);
        cv::waitKey(1);

}

instead of the expected corners, I get this result:

What is going wrong here? Why are my corners not aligned?

I have tried other images, and i see the same thing.

Thank you!

Ah, as usual, user error. I was passing the number of squares in the format that findChessboard needs, using the inner

Hi, thanks for your reply. As in the image above, the corner points are misaligned. However, I have just figured this ou

Detecting ChAruco board using example code, strange result. I am using the following code, pulled straight from the tuto

Detecting ChAruco board using example code, strange result. I am using the following code, pulled straight from the tuto

Ah, this will do it. solvePnPGeneric instead of regular solvePnP. std::vector<cv::Vec3d> rvecsVec, tvecsVec;

Return both solutions from solvePnp cv::SOLVEPNP_IPPE_SQUARE? I have a PnP function that uses solvePnP with the cv::SOLV

Return both solutions from solvePnp cv::SOLVEPNP_IPPE_SQUARE? I have a PnP function that uses solvePnP with the cv::SOLV

Hi, thank you for your answers. I am successfully creating a point cloud, using this: float fx = 284.492615; float

Calculating depth with Ptr<stereosgbm>, black bar on image I have a calibrated fisheye stereo pair that I am using

Setting the cx and cy values manually to the real middle of the image size has helped me here. Creating the Projection

cv::fisheye::stereoRectify giving black image. I am trying to create a depth map from a fisheye stereo camera. I have in

cv::fisheye::stereoRectify giving black image. I am trying to create a depth map from a fisheye stereo camera. I have in

Get 3d point from clicked depth map pixel? I have a depth map, from the Zed stereo camera. I need a function that gives

Hi, I have a calibrated stereo camera, so i have the extrinsics, intrinsics and dist cooeffs.

I have a chessboard, and using :

bool findChessboardCornersAndDraw(Mat inputLeft, Mat inputRight) {
    _leftOri = inputLeft;
    _rightOri = inputRight;
    bool foundLeft = false, foundRight = false;
    cvtColor(inputLeft, inputLeft, COLOR_BGR2GRAY);
    cvtColor(inputRight, inputRight, COLOR_BGR2GRAY);
    foundLeft = findChessboardCorners(inputLeft, boardSize, cornersLeft, CALIB_CB_ADAPTIVE_THRESH | CALIB_CB_NORMALIZE_IMAGE);
    foundRight = findChessboardCorners(inputRight, boardSize, cornersRight, CALIB_CB_ADAPTIVE_THRESH | CALIB_CB_NORMALIZE_IMAGE);
    drawChessboardCorners(_leftOri, boardSize, cornersLeft, foundLeft);
    drawChessboardCorners(_rightOri, boardSize, cornersRight, foundRight);
    _leftOri = displayMode(_leftOri);
    _rightOri = displayMode(_rightOri);
    if (foundLeft && foundRight) {
        return true;
    }
    else {
        return false;
    }
}

I get the chessboard points. What i need to do now, is write a function that returns the 3d triangulated coordinates of those points, in relation to the camera.

Is there a function existing that will do this for me? Should I use:

cv::triangulatePoints(cam0,cam1,cam0pnts,cam1pnts,pnts3D);

Thank you!

test transform between two cameras with projectPoints? I have a stereo calibrated camera pair, so i have known intrinsic

I am trying to replace the following function with opencv: (from: https://github.com/stevenlovegrove/Pa...)

  uint64_t loadDepthFromPath(std::string filepath, pangolin::Image<unsigned short> & depth)
    {

        pangolin::TypedImage depthRaw = pangolin::LoadImage(filepath, pangolin::ImageFileTypePng);

        pangolin::Image<unsigned short> depthRaw16((unsigned short*)depthRaw.ptr, depthRaw.w, depthRaw.h, depthRaw.w * sizeof(unsigned short));

        for (unsigned int i = 0; i < 480; i++)
        {
            for (unsigned int j = 0; j < 640; j++)
            {
                depth.RowPtr(i)[j] = depthRaw16(j, i) / 5;
            }
        }

        uint64_t time = getCurrTime();
        return time;
    }

This returns a pangolin::Image<unsigned short>, that is then passed to a function:

initModel(depth.ptr);

where depth.ptr is unsigned short *

I am trying to remove pangolin from this, and use opencv instead.

So I load the same image:

     std::string frame1 = "rgbd_dataset_freiburg1_desk/depth/1305031453.374112.png";

Mat img1;
img1 = imread(frame1, 0);

convert it to unsigned short:

    cv::Mat mat_ushort1(img1.cols,img1.rows, CV_16UC1);
img1.convertTo(mat_ushort1, CV_16UC1);

Then use the pointer to the data to pass to the function:

initModel((unsigned short*)mat_ushort1.data);

The problem is, I see a very different result when I use this method. Where am i going wrong here? How can I replicate the pangolin function with openCv?

thank you.

Hi, using new images did indeed help, but I ended up going with a stereo calibration type approach, adding bundle adjust

Thank you! I cannot sync them, it's true. I will try to use a static chessboard, and adjust the focus / lighting conditi

Calculate extrinsics between two cameras using FindHomography? I have two cameras with different lenses and resolutions.

Hi, I just noticed an error in the code I uploaded, the read should be: cv::Mat frame1_2 = cv::imread("frames/frame

Sure, you can grab the intrinsics and frames here: https://drive.google.com/file/d/1GJo_D4MTiAKlYU3kQduHIbhG2klg1j_S/vi

Calculate extrinsics between two cameras using FindHomography? I have two cameras with different lenses and resolutions.

Calculate extrinsics between two cameras using FindHomography? I have two cameras with different lenses and resolutions.

Calculate extrinsics between two cameras using FindHomography? I have two cameras with different lenses and resolutions.