Mpeg-4 Video Decoding Platform-based Design And Realization

MPEG-4 has the characteristics of effective compression, adaptive bit rate, fine fault-tolerance and interaction based on contents. So it has good applied and developmental outlook. But the stream encoding and decoding technology of MPEG-4 makes use of theories and technologies on signal processing, networks and computer synthetically. Thus it is difficult to study or develop it. The task of this research is to design and implement a video decoding platform based on MPEG-4 in order to enhance developing quality and shorten the developing period for the programmer. The programmer can design a stream media system based on MPEG-4 conveniently and effectively by calling the interface supported by the platform.This thesis begins with a brief analysis of the current video compression standards. Based on the analysis, the comprehensive functional module planning and specifications of calling interface of the platform are designed. Then the key technologies and implementation method are introduced with emphasis. The decoding theory and implementation method of texture, shape and moving information in the picture are given in detail. The platform adopts repetitive padding and decoding based on macro block and context-based arithmetic decoding techniques to solve the problem of VOP with any shapes. This thesis introduces the implementation of Fine Granular Spatially Scalable decoding to meet the demands of networks application in detail. In order to guarantee the play quality of the receiver, a double-buffer structure is proposed to solve the jitter problem brought by networks latency. And the measures took in the platform to compensate package lost and conceal errors are proposed. In the end, an application example developed by using the decoding platform is given.