| HEVC(High Efficiency Video Coding) is the newest international video coding standard after H.264,It was developed by the Joint Collaborative Team on Video Coding(JCT-VC) consisting of experts from ITU-T’s Video Coding Experts Group (VCEG) and Moving Picture Experts Group(MPEG). At the same picture quality,the compression efficiency of HEVC has more than double, compared with H.264. However, a high coding compression is at the expense of computational complexity, computational complexity of HEVC is about 2 to 3 times of H.264, Then it is necessary to optimize its algorithms for the implement in real-time applications.1) In reference software HM-10.0, integer pixel motion estimation fast algorithm search points by sequentially steps 1,2,4,8,16,32,64 in the search center every time. According to the experimental statistical conclusions,the probability of the best point found in the search center and steps 1 and 2 is maximum,this paper sets threshold to stop early search and obtain the rule of the thresholds,that is, almost still movement,gentle movemen and dramatic movemen sequences use different threshold to stop early search.2) Traditional methods obtain the optimal solution with function,then a fine search is used to obtain the best match poit near optimal solution.However,a fine search increases search points and introduces the computational complexity. In this paper,a fast fractional-pixel motion estimation search algorithm is proposed. It uses two quadratic functions which project an ellipse to obtain two fractional-pixel candidate points,the best point is obtained by weighting these points.Experimental results show that the proposed method can saves about 47.44% with negligible quality degradation.3) At last, this paper combines the two proposed motion estimation fast algorithm, and the experimental results show that the combined algorithm saves averagely about 58.56% of the encoding time with negligible loss of the video quality and a little decrease of the bitrate. |