A true motion estimation method based on binarized cross correlation

Francesco Michielin, Giancarlo Calvagno, Piergiorgio Sartor, Oliver Erdler

Abstract: In this paper a new method that combines the traditional approaches of the phase correlation and block based recursive search for motion estimation is presented. Phase correlation is generally more complex but provides an immediate convergence and it is less affected by noise. Recursive search is already suitable for real time applications thanks to its wide utilization, but the increase of resolution in modern displays and the need for real time software implementations demand for faster approaches. The proposed method adresses this problem by means of a binarized cross correlation as a matching cost criteria for a recursive search strategy. The results show a similar vector field estimation in comparison to the standard SAD based recursive search method with a considerable reduction of the computational effort.

External IDs:dblp:conf/icce-berlin/MichielinCSE13