Experimental results for image-based pose and velocity estimation

A method is presented to estimate the pose and velocity of a moving object in Euclidean space. Euclidean homography relationships are used to recover the pose of an object with respect to a camera frame. A Lyapunov-based robust integral of the signum of the error (RISE) control law is employed to estimate the velocity of an object with respect to a camera frame. The method described in this paper does not require the initial rotation between the camera frame and the object reference frame but requires only a single known length on the object. It is distinguished from earlier methods which require rotation matrix along with a single known length. A series of experiments are presented to test the accuracy of the algorithm. The pose estimates are calculated and compared with ground truth data logged using a laser speed gun - a velocity measuring device. Velocity estimates are compared with ground truth data at various speeds.