The Mpeg-2 To H.264 Transcoding Algorithm Research

Along with digital AV and multimedia technology developing, the international video coding standards, such as H.261, MPEG-1, MPEG-2, H.263, MPEG-4, H.264, have been established and applied to different applicaton areas. As there existed many kinds of video coding standards, the interoperability between different video equipments is becoming more and more important. Video transcoding is one of the key enabling technologies for multimedia interoperability.MPEG-2 is the most common video coding standards adopted in the multimedia industry. Compared with the previous video coding standards, under the same video quality, H.264 has better encoding efficiency as well as better internet adaptability and error robustness. Therefore it is considered wildly as the next-generation video coding standards. In order to adapt different application environment, the transcoding between MPEG-2 and H.264 is a vital issue. This thesis focuses attention on the research of video transcoding.This thesis has introduced the goal and the development of video transcoding, provided a brief overview of the structure and the flow of the H.264, especially introduced the key technologies of VCL used by H.264. Based on analysing the similarities and differences between MPEG-2 and H.264 and studing the existed video transcoding architectures, then some research has been done on the implementation of transcoding of I frames in DCT-domain, the implementation of open-loop transcoding algorithm in DCT-domain, and the optimization of CPDT.After analysing the theories of converting the 8×8 DCT blocks into H.264 4×4 DCT blocks in DCT-domain and intra-frame prediction in pixel-domain, this thesis has introduced the implementation of transcoding I frames in DCT-domain in detail and analysed the computational complexity.During transcoding MPEG-2 to H.264, first discussed the open-loop transcoding algorithm in DCT-domain. Converting the 8×8 DCT blocks into H.264 4×4 DCT blocks in DCT-domain can reduce the computation of H.264 integral DCT transform. Second introduced a method to optimize the CPDT algorithm, using the energy of residual MB of MPEG-2 to determine the block mode can achieve similar block proportion with the method of checking all block mode. By reusing the motion information from MPEG-2, we can set a small search range. Simulation results demonstrate that the optimizated algorithm can achieve similar image quality with CPDT.