| One of the development trend in video coding technology is the pursuit of higher coding efficiency. The H.264standard has been a great success in terms of both coding efficiency enhancement and flexibility for effective use over a broad variety of network and application domains. However, an increasing diversity of services, the growing popularity of HD video, and the emergence of beyond-HD formats (e.g.4k×2k or8k×4k resolution) are creating even stronger needs for coding efficiency superior to H.264’s capabilities. Under this circumstance, MPEG and VCEG have established a Joint Collaborative Team on Video Coding (JCT-VC) to develop a successor to H.264. This new international standard is referred to as High Efficiency Video Coding (HEVC). Although HEVC can achieve up to approximately50%bit rate savings for similar perceived video quality compared to H.264, the wide and deep penetration of H.264creates a need for the co-existence of these technologies in a fairly long period of time. Therefore, transcoding from H.264to HEVC has important practical significance in the network transmission and storage. However, on one hand, HEVC introduces a series of time-consuming coding algorithm in order to improve the coding efficiency, thus brings new challenges to real-time video transcoding applications. Considering the characteristics of HEVC coding algorithm, making full use of the H.264bitstream information to speed up the HEVC re-encoding process of transcoding is one of the key technologies to improve the performance of transcoder. On the other hand, since the goal of transcoding is to obtain higher compression efficiency at the same video quality and the analysis of video saliency in computer vision and image processing has become an important research field, how to extract the video saliency from the compress domain of H.264bitstream from a human visual perception to guide the optimization of the transcoding from H.264to HEVC is also one of the key technologies to improve the performance of transcoder. In this context, the research of video transcoding from H.264to HEVC is carried out.In chapter1, the importance of the research work is firstly presented. Secondly, the video codec technology, H.264and HEVC codec technology, video transcoding technology and video saliency and its applications are biefly summarized. Finally, the main research contents and the structure of the thesis are illustrated.In chapter2, low-complexity video transcoding algorithm from H.264to HEVC is studied based on statistical analysis. For inter frame transcoding, by exploiting the mapping correlations of Skip mode in HEVC and all modes in H.264/AVC, an early decision of Skip mode is firstly introduced. Secondly, through the statistical analysis of coding bits, a fast prediction unit (PU) partition selection for both symmetric and asymmetric partitions is described. Finally, the motion estimation process is optimized according to the motion similarity between H.264and HEVC bitstreams. For intra frame transcoding, the searching depth range of coding tree unit (CTU) is firstly decided based on the the intra coding modes in H.264. Secondly, gradient directions are statistically calculated and a gradient-mode histogram is generated for each coding unit. Finally, based on the distribution of the histogram, only a small part of the candidate modes are chosen for the intra coding processes.In chapter3, a fast H.264to HEVC transcoding algorithm based on region feature analysis is proposed. First, each frame is segmented into three regions in units of CTU based on the correlation between image coding complexities and coding bits of the H.264source stream. Then the searching depth range of each CTU is adaptively decided according to the region type. After that, motion vectors are de-noise filtered and clustered in order to analyze the region features of coding unit (CU). Based on the analysis results, the minimum searching depth of CU and partitions of PU are optimally selected, and the motion vector predictor and search window size of motion estimation are also optimally decided for further reduction of the computational complexity.In chapter4, based on the characteristics of human vision system, a video saliency based video transcoding algorithm from H.264to HEVC is proposed. Firstly, the motion vector field included in the H.264bitstream is utilized to do the global motion estimation and motion segmentation. Secondly, the distribution of coding bits is combined to produce the visual saliency map which indicates the salient regions in the videos. Finally, during the re-encoding process, a frequency coefficient suppression technique in the transform domain is used in the non salient region for further bitrate reduction while keeping the subjective quality of salient region.In the final chapter, the novel achievements of the research in this thesis and the prospect of the future research are concluded. |
PDF Full Text Request