Adaptive asymmetric diamond search algorithm for block-based motion estimation

Abstract

We present an adaptive asymmetric diamond search (AADS) for block based motion estimation. The proposed method exploits an adaptive search pattern based on the predicted motion vector. In addition, the accuracy of the prediction is improved by changing the initial search center position. Simulation results show that the proposed algorithm can reduce the number of search points by up to about 50% compared to the regular diamond search while keeping the picture quality similar to other fast search algorithms.