An adaptive search range decision algorithm for fast motion estimation

초록

In this paper, we present an adaptive motion search range decision algorithm for low-power video encoder. The performance, computation, and power consumption of block matching motion estimation algorithms in video coding standards depends on the motion search range. The motion search range to restrict the motion candidates controls the trade-off between the motion accuracy and the encoder complexity. The proposed algorithm adaptively determines the motion search range by local statistics of the neighbor blocks, resulting in dramatic reduction of computational cost of video encoder without the loss of coding efficiency. Experimental results show that the proposed algorithm speeds up encoding time by 1.5-2.8 times, and reduces power consumption to 2.5-7.0 times.