Why do real world object points and corner points not match?

Hello everyone,

I am working on camera calibration. With the help of the cameraCalibrate function in OpenCV, we can find the intrinsic and extrinsic parameters of the camera. This function takes the object’s real point and the pixel equivalent coordinates of these points as input.

It is stated that while OpenCV determines the coordinates of the object in the real world, it takes the top left corner as a reference. Therefore, the upper left corner is at the coordinates (0, 0, 0). But, when I find the corners with findChessboardCorners function in OpenCV, it brings the bottom left coordinate of the object as the first corner. Assume the distance between two squares is 30 mm. However, according to the real world, the equivalent of this part is normally expected to correspond to the position (150,0,0).

How opencv handle this situation ?


Thanks for your valuable comments.