Using the Homography matrix between two images (Target image position and Current image position), we decompose homography in order to take the rotation and translation. So the P3-DX use that directions in order to go in the target position

PeopleBot track a tennis ball using the Mean-Shift Algorithm which is an efficient approach to tracking objects whose appearance is defined by histograms. Then use the distance of the object from the center of the video plane in order to move.

StereoBot track an object using the relative depth of the stereo theory in order to move.