Question about the extrinsic calibration homography #85
Description
Hi
I have a question about the homography obtained via the extrinsic_calibration procesure, since I cannot figure out if it is working correctly or if I have understood it properly.
I would have guessed that this matrix represents the transformation between a pixel (u, v, 1) and a point on the ground (x, y, 0) on a frame attached to the robot.
If so, I would expect that providing the 4 image corners I would get 4 points which correspond to the 4 corners of the ground that the camera can see. What I get are instead negative values for both x and y (I am not sure of which is which but one of them should always be positive?). Have I misunderstood something?
Maybe a related question, how does this transformation relate to the transformation that I have to apply to the image to get the birds-eye-view of the ground?
Thanks in advance for any help