Inaki's profile - activity

Hi!

I have some problems to apply the Kalman filter to a ROI of a program I made.

I am very interest in Kalman filter, because I want to reduce the noise when i am tracking the ROI.

I started using the program published by Roy shown in http://www.morethantechnical.com/2011/06/17/simple-kalman-filter-for-tracking-using-opencv-2-2-w-code/

In the program the Kalman filter (2D) is used to track the mouse.

I am trying to adapt the code for tracking a ROI. My ROI is defined:

//ROI

lado = 2*R; // R is the radius of a previously defined circle

origen_x = (int) ( p[0] - lado/2.); // p[0] and p[1] are the coordinates of the center of the circle

origen_y = (int) ( p[1] - lado/2.);

I've tried to apply the Kalman filter this ROI:

int main (int argc, char * const argv[]) {

Mat img(500, 500, CV_8UC3)

KalmanFilter KF(4, 2, 0);
Mat_<float> state(4, 1); // (x, y, Vx, Vy) 
Mat processNoise(4, 1, CV_32F);
Mat_<float> measurement(2,1); measurement.setTo(Scalar(0));
char code = (char)-1;

for(;;)

namedWindow("kalman");

             KF.statePre.at<float>(0) = origen_x;  /the x coordinate of my ROI
    KF.statePre.at<float>(1) = origen_y; // the y coordinate of my ROI
    KF.statePre.at<float>(2) = 0; //No info of the velocity
    KF.statePre.at<float>(3) = 0; 
    KF.transitionMatrix = *(Mat_<float>(4, 4) << 1,0,0,0,   0,1,0,0,  0,0,1,0,  0,0,0,1);

    setIdentity(KF.measurementMatrix);
    setIdentity(KF.processNoiseCov, Scalar::all(1e-4));
    setIdentity(KF.measurementNoiseCov, Scalar::all(1e-1));
    setIdentity(KF.errorCovPost, Scalar::all(.1));

//randn(KF.statePost, Scalar::all(0), Scalar::all(0.1));

    for(;;)
    {

// Point statePt(state(0),state(1));

        Mat prediction = KF.predict();
        Point predictPt(prediction.at<float>(0),prediction.at<float>(1));

                      measurement(0) = origen_x;
        measurement(1) = origen_y;

        Point measPt(measurement(0),measurement(1));
                    // generate measurement
        //measurement += KF.measurementMatrix*state;

        Mat estimated = KF.correct(measurement);
        Point statePt(estimated.at<float>(0),estimated.at<float>(1));
        kalmanv.push_back(statePt);

        // plot points

#define drawCross( center, color, d ) \ line( img, Point( center.x - d, center.y - d ), \ Point( center.x + d, center.y + d ), color, 2, CV_AA, 0); \ line( img, Point( center.x + d, center.y - d ), \ Point( center.x - d, center.y + d ), color, 2, CV_AA, 0 )

        img = Scalar::all(0);
        drawCross( statePt, Scalar(255,255,255), 5 );
        drawCross( measPt, Scalar(0,0,255), 5 );

// drawCross( predictPt, Scalar(0,255,0), 3 ); // line( img, statePt, measPt, Scalar(0,0,255), 3, CV_AA, 0 ); // line( img, statePt, predictPt, Scalar(0,255,255), 3, CV_AA, 0 );

// randn( processNoise, Scalar(0), Scalar::all(sqrt(KF.processNoiseCov.at<float>(0, 0)))); // state = KF.transitionMatrix*state + processNoise;

        imshow( "kalman", img );
        code = (char)waitKey(100);

        if( code > 0 )
            break;
    }

i I know that is not complete, but I need some ideas and help to continue (I repeat that the program without the Kalman Filter works ok, I can detect and track a ROI, but I need to implement this filter to reduce the noise).

Thaks ... (more)