I have trained a cascade classifier to detect an object that is multi-form. I now want to know the exact position of the in the frame and just move my step motors to follow the object. I have splitted the image into four equal images , each image corresponds to one camera auto-calibration movement movement. I just want to move my camera automatically in case the object detected is too much on the upper left ,upper right etc and put it at the center of the field of view