Research On High-Efficiency Intra Prediction Algorithm Based On Video Texture Analysis

With the fast growth of the mobile communication, the developing direction of the multimedia application technologies tends to be networked, interactive and realistic. Among them, video has drawn many attentions because of its outstanding performance, such as abundant content, vivid and direct viewing effect and so on. On the other hand, the coding effect of video coding protocol also has a qualitative leap through experts and scholars’ continuous research and development. By introducing some novel technologies such as various demensions of prediction blocks and prediction modes, and rate distortion optimization (RDO), H.264/AVC which is the dominant video coding protocol achieves more than 50%coding performance (such as coding quality and bit rate controling) optimization over H.263. Moreover, the latest generation of video coding standard HEVC further expands the coding block size and its prediction mode on the basis of H.264/AVC to reduce the output bit rate of 50% to further reduce the transmission pressure under the same coding quality. The high quality of coding performance, however, has led to the encoding complexity by soaring exponential, makes the general equipments, especially for the popular mobile intelligent device over the future wireless network environments cannot complete real-time encoding. Therefore, the research on reducing the computational complexity as much as possible while maintaining the video quality and compression efficiency become most important issue in the video coding standard.It gives a brief introduction on the history of video coding standards and its technical framework at the beginning of this thesis. Then, the key technologies of H.264/AVC and HEVC are deeply studied, especially for intra-picture prediction process which leads to soaring coding complexity. It has laid a solid foundation for proposing high quality low complexity intra-picture prediction algorithm.In order to reduce the computational complexity of the H.264/AVC intra coding, a high-efficiency mode decision procedure based on energy distribution for H.264/AVC is proposed in this paper. First, a fast intramode-type selection algorithm is presented to determine the intraprediction block size in advance. Then, the original multiple mode candidates for each prediction type are reduced at least by half with the information abstracted from macroblock and its 4x4 subblocks. The simulation results show that the proposed algorithm could reduce around 84% encoding time in average while maintaining the encoding performance efficiently.To alleviate the computation load of HEVC intra process, this paper also proposes a fast mode decision algorithm based on texture complexity and direction for HEVC intra prediction. Firstly, a fast Coding Unit (CU) selection algorithm according to each depth levels’texture complexity is presented to filter out unnecessary prediction units. And then, the original redundant mode candidates for each prediction unit are reduced according to its texture direction. The simulation results show that the proposed algorithm could reduce around 56% encoding time in average while maintaining the encoding performance efficiently with only 1.0% increase in BD-rate compared to the test model HM10.0 of HEVC.The proposed fast intra prediction algorithm for H.264/AVC and HEVC standard make full use of the characteristics of the video content itself and make correct pre-judgment for the mode decision of intra coding. Moreover, it has obtained an ideal coding effect. Therefore, the proposed algorithm in this thesis provides a certain reference value for the research of video coding and providing satisfied effect for the users in mobile communication environments.