Research On Intra Coding Optimization Algorithm For 3D-HEVC

In recent years,the rapid development of photography technology has led to the rapid development of HD and ultra-high-definition video.Although the corresponding hardware processing equipment has some improvements,video compression coding has been an important research issue.Compared to the previous H.246 encoding standard,the latest international video encoding standard HE VC can reduce the compression rate of high-resolution video by half in the case of coding quality remains unchanged.While HD video development to a certain stage,it began to change to three-dimensional video.The research of multi-view naked eye 3D video and 360-degree panoramic video is in the ascendant.Panoramic video can give users a better experience of immersion,but the information of panoramic video is much larger than high-definition video.Thus,the coding advantages of HEVC can be further reflected in the encoding of panoramic video.This thesis focuses on the research of 3D video coding technology(3D-HEVC)based on the HEVC standard.Compared with ordinary high-definition video,the principle and process of 3D-HEVC encoding is more complicated,the encoding framework of 3D-HEVC is more cumbersome,the time consumed by 3D-HEVC encoding is also greatly increased.Therefore,in view of the above problems,this thesis mainly includes:(1)The human visual system has various concealment effects and shielding effects,and the human eyes can only detect the signal changes with more than a specific threshold.Therefore,this thesis introduces the visual feature into the 3D-HEVC encoding technology,and proposes an optimized video encoding algorithm based on the pixel-domain Just Noticeable Distortion(JND)model.This method replaces the complex rate-distortion optimization process by introducing the judgment of JND threshold which is used for fast division of Coding Unit(CU).This method effectively reduces the complexity of encoding and saves the encoding time.;(2)Because the total bits of the coding unit are closely related to the texture complexity of the video itself,the intra-frame optimization algorithm is proposed in this thesis which not only takes into account the visual characteristics of the human eyes,but also introduces the total bits of the coding unit as a criterion for the division of coding units.According to these two characteristics,we can find the optimal JND and the total bit threshold to optimize the coding unit division by training the two characteristics.This method can directly determine the corresponding threshold of division and save a lot of statistic time which effectively reduces the encoding time under the premise of guaranteeing the quality of coding reconstruction;(3)This thesis researches the generating principle of panorama and panorama video,and analyzes the cylindrical stitching algorithm of panorama.Because the panoramic video information is much larger than the amount of information contained in the high-definition video.It is not efficiency by using H.264 encoding algorithm to encoding.Due to the complexity of the data structure,the compression coding of panorama has not yet formed a unified standard.In this thesis,the HEVC standard algorithm is applied to the encoding of panoramic video,and a HEVC-based panoramic video coding algorithm is proposed,which effectively improves the rate-distortion performance of the encoding.