Confidence interval based motion estimation

A new video standard called High Efficiency Video Coding (HEVC) is now being finalized. In comparison with the H.264/AVC video coding standard, HEVC further improves video coding rate distortion (RD) performance, but at the price of significant increase in its encoding complexity, especially in its motion estimation (ME). To reduce the ME complexity in HEVC while maintaining its RD performance, in this paper, we first formulate ME as a statistical inference problem and then propose a confidence interval based ME method. It is shown by experiments that, for the four test sequences with higher searching complexity under low delay main, our proposed ME method further reduces the integer level ME time of the fast search in HEVC by 73.49% on average with only 1.22% increase in bit rate and 0.024dB loss in PSNR.