The Analysis Of Syntax Elements And Code Rate Estimation Algorithm Based On AVS2 Standard

In today’s society, multimedia information has become the main body in the process of information transmission, its rapid development and increasing user needs make video compression technology constantly updated and improved. In foreign countries, in the case of the emergence of H.264/AVC and the High Efficiency Video Coding Standard for Next Generation--H.265/HEVC, our country also vigorously develop with independent intellectual property rights of the video coding standard, launched the "Audio Video coding Standard (AVS) and the radio and television advanced audio and video coding and decoding part 1:video" (AVS+). Currently, the next generation of AVS+which named AVS2 is in the process of standardization.AVS2 video coding syntax element is the structure foundation of realization and interpretation of entropy code stream. Because of AVS2 reference code RD7.0-RD11.0 (2014.Jun-2015.Jul) Documentation about the syntax elements which is incomplete and there are lots of inconsistencies with the actual code. Thesis makes a detailed analysis of all the syntax element and the code stream structure under all-intra configuration conditions and gives the analysis report based on the actual code RD11.0, which has important reference value for study and improve of AVS2 video coding algorithm.Thesis studies the rate-distortion optimization techniques in mode selection about AVS2 video encoding structure, which found that calculating the minimum cost to choose the best candidate model in all the candidate patterns must calculate the rate for many times, and rate calculation must be achieved by entropy coding. Due to repeated calls to a huge number of entropy coding, it takes a lot of time. This paper proposes an approximate rate estimation algorithm instead of entropy coding to get rate in the optimum prediction mode selection process. This method calculates and estimates approximate rate by doing statistics of the maximum probability and minimum probability about binary symbols to be encoded instead to obtain the true rate by calculating the actual entropy coding. A variety of YUV video tested in RD11.0 encoder, the encoding test results showed that the time of rate distortion optimization selection rate calculation process reduced by an average 69.49%, and the time of total video encoding reduced by an average 10.99%,but the rate increase not more than 1.5%, PSNR has almost no difference with standard codes. This paper is divided into four chapters:The first chapter:This chapter introduces the background and research status of video encoding, as well as related research and comparison of the status quo at home and abroad. Aimed to the issue of improve the compression efficiency but dramatically increased the complexity of encoding time, this chapter elaborated it.The second chapter:This chapter introduces AVS2 video coding coding framework, intra & inter prediction, transform quantization, entropy coding, and makes an exposition about rate-distortion optimization technique.The third chapter:The rate of AVS2 structure and its syntax elements were analyzed.The forth chapter:for mode selection in the rate-distortion optimization technique, a rate estimated algorithm was proposed which replaced rate calculating by entropy coding, The experimental results and analysis.