A Study of Fast High Efficiency Video Coding (HEVC) Encoder Using Tempo-Spatial Correlation

• Main Authors: Wan-Ying Jhuang, 莊宛穎
• Other Authors: Chou-Chen Wang
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/40444731443441565547

碩士 === 義守大學 === 電子工程學系 === 101 === With the rapid development of electronic technology, the high-resolution panels of 4K*2K (or 8K*4K) will become the main specification of large size digital TV in the future. However, the current H.264 video coding standard can’t support the video applications of ultra-high definition (UHD) resolution. Therefore, the JCT-VC began to develop a new high efficiency video coding (HEVC) for video compression standard to satisfy the UHD requirement in January 2010, and the first version of HEVC was finalized by JCT-VC in January 2013. HEVC can achieve 50% bit rate reduction when compared with the H.264 under the same video quality. Because of the HEVC adopts some new coding techniques including coding unit (CU), prediction unit (PU) and transform unit (TU). The CU can be split by coding quad-tree structure of 4 level depths (64*64 to 8*8) for inter/intra prediction. The PU is used to for performing the related to the prediction processes. When pruning the best CU coding quad-tree, the inter prediction module executes 7 different prediction modes including SKIP mode、intra2N×2N、intraN×N、inter2N×2N、inter2N×N、interN×2N and interN×N to find the best mode. Especially, in the inter2N×2N、inter2N×N、interN×2N and interN×N prediction need perform motion estimation (ME) and motion compensation (MC). However, ME process is performed using all the possible depth levels and prediction modes. HEVC can achieve the highest coding efficiency, but requires a very high computational complexity such that it is difficult to reach real-time applications. In order to reduce the computational complexity of HEVC encoder, there are many fast encoding methods to speed up the PU modules. Recently, Liu et al. proposed an adaptive CU depth range determination (ACUDRD) algorithm based on the maximal and minimal values of depth levels to determine current CU depth. However, the ACUDRD is ineffective for video sequences with active motion or rich texture. Therefore, in order to further improve the performance of ACUDRD algorithm, we utilize the characteristics of strong temporal and spatial correlation in natural video sequences. Firstly, we analyze the temporal correlation and find that the probability is near to Gaussian distribution. And then, the depth of current CU is predicted by 9 Gaussian weighting from neighboring blocks of the co-located block. On the other hand, we also analyze the spatial correlation and predict the depth of current CU using 4 best weighting from encoded neighboring blocks. Finally, we determine the best predicted depth range of current CU from the intersection of two predicted depth range. The simulation results show that the proposed algorithm can reduce the processing times of PU about 31% ~ 66% and achieve an average time improving ratio (TIR) about 37% ~ 72% when compared to HEVC(HM8.1). Compared with ACUDRD algorithm, the proposed algorithm can further reduce the processing times of PU about 3% ~17% and further achieve an average time improving ratio (TIR) about 7% ~ 18%. It is clear that the proposed algorithm can efficiently reduce the computational complexity of HEVC encoder with insignificant loss of image quality.