I have all the coordinate frame (by opencv extraction), and how all the pixels in the original image is obtained by the coordinate information? (I'm not familiar with opencv),