Research On Low Complexity And Low Power Consumption Avs High Definition Video Encoders

With the development of multimedia technology,the demand of video resources is improved,which brings a great challenge to video storage and transmission technology.After years of development,the international video coding standards have experienced a number of generations of evolution,which greatly promote the development of video industry.However,the high cost of the international video coding standard has caused great obstacles to the development of China's video industry.In order to solve the monopoly problem of video coding standards,China established the Audio and Video coding standard workgroup(AVS group),and proposed the first generation of video coding standard,AVS,in 2003.AVS achieved the equivalent coding efficiency with H.264/AVC and has become an international standard.With the increasing demand for high-definition,ultra high-definition and 3D videos,the second generation of video coding standard AVS2 is proposed and is about to be completed.The collection of a large number of new video coding technologies greatly improve the efficiency of video coding,AVS series standards have a broad development prospects and industrial basis in our country.But in the practical application,the coding complexity is still a serious problem.In this paper,we research and explore the problem from the actual needs and deeply look into the optimization potential of the algorithm and platform,improve the high-definition video coding efficiency on the low power consumption platform from multi-levels.In this paper,the development history of the video coding standards is briefly reviewed,and the key problems in the industrial process of AVS series standards are also analyzed based on the video industry needs.Under the background of the fast development of the multimedia equipment,we start from the optimization of AVS series standards on low power consumption platform,which is developing rapidly,and study the complexity problem in the needs of high definition video and low power application.On the basis of comparison and analysis of the AVS and AVS2 video coding standards,this paper introduces the main modules and key technologies of the two standards and optimize the AVS series standards from different levels with the combination of current research status and application requirements.The optimization of video coding can be mainly divided into two aspects: the algorithm and the platform.In this paper,we deeply research the relationship between the complexity corresponding coding parameters and the coding video sequences,and establish a video texture characteristics based coding parameters control mode to optimize the complexity of coding process.Starting from the practical application of low power consumption platform occasions,we realize a high quality and fast speed coding process.Combining with the platform characteristics,this paper optimize AVS from instruction parallel and thread parallel.With the help of instruction sets and thread parallel interface,we realize decades times speed promotion with less than 30 W power consumption.This paper further study the latest research results of AVS group,AVS2,from the point of algorithm,and proposed a texture information based fast mode and depth decision algorithm in view of the high complexity intra frame coding.With the gradient based texture analysis,the prediction modes are predefined and selected to reduce the computing process in mode selection.And according to the recursive coding units partition process,we introduce a texture variable to classify the coding units and proposed a bottom up partition mode to skip the unnecessary computation.Experiments show that this algorithm can reduce nearly 50% coding time with less than 0.05 dB PSNR loss.In summary,the AVS encoders have excellent development foundation and broad development prospects.The optimization of AVS and AVS2 encoders is of important research and application value.