Optical Flow visualization

Question

I am trying to visualize the output of calcOpticalFlowPyrLK() (OpenCv v3.0.0). I am not trying to draw whole image with optical flow, only the direction arrow. The problem is, I can't get to the output as in the examples. Every 10 frames I renew the points for the calculation of the flow. The function itself

calcOpticalFlowPyrLK(CentroidFrOld, CentroidFrNow, mc, CornersCentroidNow, feat_found, feat_errors, Size(15, 15), 2, cvTermCriteria(CV_TERMCRIT_ITER | CV_TERMCRIT_EPS, 10, 0.03), 0);

Where CentroidFrOld is grayscale frame, CentroidFrNow is grayscale frame+1, mc is a vector<Point2f> array of points and CornersCentroidNow is an empty array waiting to be filled with new points.

When drawing them I use simple code:

for (size_t i = 0; i < CornersCentroidNow.size(); i++){             
    if (feat_errors[i] > MAX_ERR || feat_found[i] == 0) continue; 
    Point p0(ceil(mc[i].x), ceil(mc[i].y)); // are the points of interest (centroids of contours)
    Point p1(ceil(CornersCentroidNow[i].x), ceil(CornersCentroidNow[i].y));     
    arrowedLine(empty, p0, p1, Scalar(0, 0, 255), 2, 8, 0, 0.2);
}

after this block of code. When I draw them every frame I get this output:

If I update the previous frame used for calcOpticalFlowPyrLK() function

CentroidFrOld = CentroidFrNow.clone();

I get this output (the line is short and it is moving foward every 10 frames - as set to get new points)

If the previous points happen to be next points as well

CentroidFrOld = CentroidFrNow.clone();
mc = CornersCentroidNow;

I get this output (the line is short, but it is moving along with the object)

The desired output I can't achieve is

Do I need to manually lengthen the line? Noone is doing so in similar examples of implemantation of Optical Flow

Answer 1

void drawOptFlowMapF(const Mat& flow, Mat& cflowmap, int step, const Scalar& color) {
    for (int y = 0; y < cflowmap.rows; y += step)
        for (int x = 0; x < cflowmap.cols; x += step)
        {
            const Point2f& fxy = flow.at< Point2f>(y, x);
            line(cflowmap, Point(x, y), Point(cvRound(x + fxy.x), cvRound(y + fxy.y)),
                color);
            circle(cflowmap, Point(cvRound(x + fxy.x), cvRound(y + fxy.y)), 1, color, -1);
        }
}
void displayF(Mat flow)
{
    //extraxt x and y channels
    cv::Mat xy[2]; //X,Y
    cv::split(flow, xy);

    //calculate angle and magnitude
    cv::Mat magnitude, angle;
    cv::cartToPolar(xy[0], xy[1], magnitude, angle, true);

    //translate magnitude to range [0;1]
    double mag_max;
    cv::minMaxLoc(magnitude, 0, &mag_max);
    magnitude.convertTo(magnitude, -1, 1.0 / mag_max);

    //build hsv image
    cv::Mat _hsv[3], hsv;
    _hsv[0] = angle;
    _hsv[1] = cv::Mat::ones(angle.size(), CV_32F);
    _hsv[2] = magnitude;
    cv::merge(_hsv, 3, hsv);

    //convert to BGR and show
    cv::Mat bgr;//CV_32FC3 matrix
    cv::cvtColor(hsv, bgr, cv::COLOR_HSV2BGR);
    cv::imshow("optical flow", bgr);
    imwrite("c://resultOfOF.jpg", bgr);
    cv::waitKey(0);
}