| The video compression coding technology has achieved a rapid development with the computer information technology since the last century 80's. The family of MPEG video coding standard that is instituted by ISO/IEC has promoted the video communications and storage applications, and MPEG2, one of the standards, has achieved the widest implementation. As the next-generation moving picture compression standard in 21st century, H.264, also known as the 14496-10 or MPEG-4 Part10, was released by JVT (Joint Video Team) in 2003, which was constituted by MPEG (Moving Picture Experts Group) from ISO/IEC, and VCEG (Video Coding Expert Group) from ITU in 2003. It has been widely accepted and become a new international video standard. Although H.264 has achieved the improvement of compression performance, the cost for the improvement is the encoding complexity. Therefore, H.264 can not achieve a large-scale commercial application and promotion.In this paper, in order to improve the performance for the real-time video compression coding, the whole architecture of H.264 is studied, the algorithms for H.264 are analyzed materially, and find the effective method to improve the real-time performance significantly.In first chapter, we introduce the background of the research and the practical significance. The video coding theory is also introduced. In second chapter, the development of video coding standards is described. The applications of the technologies for H.264 are introduced in detail, the improvement of these technologies for the encoding performance and the impact of coding complexity are analyzed.In this paper, the coding efficiency of H.264 has been improved from the motion estimation and bitstream switching.In the study of the motion estimation, we proposed a new algorithm which can reduce the coding complexity with an adaptive motion search template selection. This algorithm can reduce the time for the motion estimation without quality degradation of the reconstruction picture.The new SP-frame is adopted in the latest video coding standard H.264/AVC. The new SP-frame, which can replace Intra-frame, is used for bitstream switching. The important characteristic of SP-frame is that the frame can be reconstructed with different reference frames. Therefore, bitstream which is coded with SP-frame can allow the bitstream switching. With bitstream switching, the server can provide the best quality of the corresponding service based on a user's network connection speed. In this paper, the SP-frame's main applications and the original SP-frame coding process are introduced; and then the new scheme is proposed, which allows drift-free bitstream switching. The proposed scheme is more efficient for bitstream switching. |
