Efficient Mode Selection Scheme for Video Coding

Abstract

Video coding plays vital role in communications and multimedia applications. The appearance and development of various new multimedia services have need for higher coding efficiency. H.264/AVC is the latest international video coding standard that can achieve considerably highest coding efficiency than previous standards and uses rate-distortion optimized (RDO) mode selection technique. This means that the encoder has to code the video by exhaustively trying all the mode combinations including the different intra and inter prediction modes. However, this mode selection process also makes the encoding process extremely complex, especially in the computation of the rate-distortion cost function, which includes the computations of the sum of squared difference (SSD) between the original and reconstructed image blocks and context-based entropy coding of the block. So, the highly complex mode selection process limits the practical implementation of codec’s. This dissertation proposes six different fast and efficient schemes to reduce the computational complexity of mode (Intra and Inter) selection process. It also can effectively reduce the spatial and temporal redundancy. In H.264/AVC intra coding, DC mode is used to predict regions with no unified direction and the predicted pixel values are same and thus smooth the varying regions which are not well de-correlated. A modified DC prediction (MDCP) mode is suggested to replace the existing DC mode based on the distance between the predicted and reference pixels. By analyzing the impact of scanning, a new Z-scan scheme is proposed which can remarkably reduce bit rates compared with AVS codec using conventional zigzag scan, while the PSNR of video sequences are maintained. The entire macroblock is divided into simple motion or complex motion contents and their cost functions are calculated by using the developed rate-based fast and efficient mode selection scheme, which plays a vital role as no additional operation module is required. To avoid the entropy coding method during intra and inter mode decision, a rate estimation method proposed and readjusted to fit the characteristic of low complexity distortion functions of SATD and SAITD that use more accurate estimation while achieving higher prediction performance with low computational requirement of feature. In addition, a relative sum of absolute difference (RSAD) based inter mode decision algorithm is proposed to determine the best inter mode without calculating xiv rate-distortion costs. To further speedup SKIP mode early termination algorithm is adopted. Based on the probability characteristics, a fast inter mode decision algorithm is suggested which can determine the best inter mode only by calculating the Hadamard Transform (H-SAD) cost instead of rate distortion cost and avoid most of the complex computation processes like quantization, variable length coding, pixel reconstruction etc. To make the motion estimation process fast, a search point reduction scheme is introduced. Then the ranking and performance of the commonly used scheme are analyzed for different kinds of video. Experimental results indicate that the proposed efficient mode selection schemes reduce the computational complexity of H.264/AVC codec without sacrificing the quality of video which can be applied in implementing real time encoder for low-power mobile devices and general purpose computers.