Adaptive Video Motion Estimation Algorithm via Estimation of Motion Length Distribution and Bayesian Classification

Real videos contain mixture of motions with slow and fast contents. No fixed fast block matching algorithm can efficiently remove temporal redundancy of video sequences with wide motion contents. In this paper, an adaptive fast block matching algorithm, called classification based adaptive search (CBAS) has been proposed. A Bayes classifier is applied to classify the motions into slow and fast categories. Accordingly, appropriate search strategy is applied for each class. The algorithm switches between different search patterns according to the content of motions within video frames. Experimental results show the proposed technique outperforms conventional standalone fast block matching methods in terms of both peak signal to noise ratio (PSNR) and computational complexity