Versatile Search Range Adjustment Algorithm for Fast Block-Based Motion Estimation

This paper proposes an adaptive search range adjustment algorithm which is particularly suitable for low-bit-depth motion estimation (ME) approaches. Typical low-bit-depth ME approaches such as one-bit transform (1BT), two-bit transform (2BT), and truncated gray-coded bit plane matching (TGCBM) perform FS with low-bit expression but these have significant PSNR drop. Combination of this proposed algorithm and low-bit expression enhance both motion estimation accuracy and time consumption. Experimental results show that low-bit-depth motion estimation approaches using the proposed algorithm perform exceptionally well in contrast to low-bit-depth motion estimation approaches without the proposed algorithm, in terms of both motion estimation accuracy and time consumption.