Frame-to-frame image motion estimation with a fuzzy logic system

Fuzzy logic systems show substantial promise for frame-to-frame motion estimation. A fuzzy logic system with simple rules is shown to provide accurate motion vector estimates with uniform and affine modeled frame-to-frame image motion of real image data. Comparisons are made with the optical flow algorithm of B. K. Horn and B. G. Schunk (1981) and the pel recursive algorithm of A. N. Netravali and J. D. Robbins (1979). These comparisons show that the fuzzy logic system presented is superior in the accuracy of its motion vector estimates. These comparisons also show that measures of picture matching such as mean squared error do not necessarily reflect the accuracy of the motion vector estimates found by an algorithm