Research On The Key Technologies In H.264/AVC Based Wireless Video Communication System

In the past decade,a rapid development of wireless communication and multimedia communication techniques has been achieved,and it resulted in great commercial success.The merge of wireless communication and multimedia techniques is accelerating recently and productions emerge continuously such as video transmission network protocols,powerful digital video signal processing chips as well as efficient video compression algorithms.The research on the video codec and video stream transmission in wireless environment has become an active task in information science and technology field.H.264/AVC is the newest international video coding standard approved by the joint video team(JVT) organization which is formed by ITU-T VCEG(Video Coding Experts Group) and ISO/IEC MPEG (Moving Picture Experts Group).H.264 adopts a number of advanced video coding technologies and has been used in all sorts of video applications because of its high coding efficiency and network friendliness.This dissertation has conducted the research focusing on the key technologies in H.264 based wireless video communication system.The contents of the task are mainly composed of fast macroblock(MB) mode selection,rate control and error resilience coding in video encoder; unequal error protection(UEP) in channel encoder;fast mode derivation of the enhancement layer MB in extended spaticial scalability(ESS) and drift resistance in fine granularity scalability(FGS) in H.264/AVC scalable video coding(SVC) extension.Our works try to decrease the computational complexity of H.264 encoder to adapt the calculation ability of mobile terminals,solve the problems of robust transmission of H.264 video over error-prone wireless networks as well as enable flexible applications over heterogeneity networks.The contents and innovations of this dissertation are as follows. First,we describe the architecture of H.264/AVC based wireless video communication system,process statistical analyses on MB/sub_MB mode in H.264 streams as well as some important encoding parameters and draw some conclusions which are foundations of our following proposed algorithms.Next,we propose a fast intra prediction algorithm (?) reduce the computational complexity of encoding intra(I) frame, (?)nd a fast MB mode decision algorithm to reduce the complexity of (?)coding predict(P) and bi-predict(B) frames.Again,we propose a solution for wireless video communication which adopts the rate control associate with the error resilience encoding and UEP motivated by the joint source and channel coding(JSCC).Finally,in SVC extension (?) H.264/AVC,we proposed an efficient method to reduce the computational complexity of the mode derivation process of the hancement-layer MBs in ESS as well as a local adaptation of leak factorαused in the leaky prediction scheme in adaptive reference FGS (AR-FGS) respectively,which both depend on the motion and texture characteristics of the co-located local MBs in the base layer.