Realization And Optimization Of BWDSP H.265 Encoder Intra Prediction

With the increasing demand for HD/UHD video in multimedia services,the previous generation of video codec standard H.264 has been difficult to meet the needs of people in video transmission and storag e.Joint Collaborative Team on Video Coding(JCT-VC)in January 2013 officially announced the latest video coding standard High Efficiency Video Coding(HEVC).As an alternative to the H.264 / AVC standard,HEVC is only about half the H.264 / AVC while maintaining the same video quality.The higher coding compression ratio enables people to have more video channels or higher-resolution video transmissions on the same infrastructure.The optimization of HEVC coding technology also brings about the improvement of algorithm complexity,it needs codec processing terminal has a higher data processing capacity in order to achieve real-time codec requirements,which has a negative impact on the promotion and application of HEVC.In this paper,a new low complexity fast intra mode selection algorithm is proposed,which can quickly determine the division of CU by analyzing the texture complexity of CU.And use the information of adjacent CU to cut the CU coarse prediction model in advance,which greatly reduces the c omputational complexity of the intra prediction process.Experiments show that the algorithm can reduce the encoding time by 45% when the coding quality is almost constant compared with that of HM-16.0? And the higher the resolution,the better the video optimization effect,which achieves the purpose of reducing the computational complexity of HEVC encoders.Based on the actual project requirements,the intra prediction module of x265 is transplanted to the BWDSP simulation platform,and the bottom computing function of the intra prediction part is optimized.Optimization work is code-level optimization,the main optimization methods are instruction parallel,loop expansion and software flow.The purpose of optimization is to make full use of the operational resources provided by BWDSP to improve program efficiency.The optimization results show that the assembly optimization work has achieved this goal.