CODEWITHSUNDEEP
c++opencvkeypoint
anti
anti

Reputation: 3145

Resizing a vector of 2d points to match a resized image?

I have an opencv application that uses blob detection to find the centers of white circles.

I run the detection, and end up with my image (at 1920 x 1080 resolution), and a std::vector<KeyPoint> of my 2d points.

Mapping the points onto the image, I get this:

enter image description here

What i need to do now is resize my image to 320 x 240, and scale the 2d points to match.

I use this code:

cv::Size imgSize(frame.cols, frame.rows); //original frame size
cv::Size resz(320, 240);                  //resized frame size

cv::Size scaleVal(imgSize.width / resz.width, imgSize.height / resz.height); //the scale value

//resize the frame

cv::resize(frame, frame, cv::Size(resz));

//Scale and draw the points:

    for (int i = 0; i < keypoints.size(); i++)
    {
        cv::drawMarker(frame, cv::Point(keypoints[i].pt.x / scaleVal.width, keypoints[i].pt.y / scaleVal.height), Scalar(0, 0, 255), cv::MARKER_CROSS);
        
    }

However this gives me:

enter image description here What am i missing here?

How can i scale a vector of points to match a resized cv::Mat image?

Upvotes: 0

Views: 462

Answers (0)

Related Questions