Video coding using the H.264 MPEG-4 AVC compression standard

Apple Computer where much of this work was done.

E-mail addresses:apuri@01a8cf0e76c66137ee061959(A.Puri),schen@

01a8cf0e76c66137ee061959(X.Chen),aluthra@01a8cf0e76c66137ee061959(A.Luthra).

0923-5965/$-see front matter r2004Elsevier B.V.All rights reserved.

doi:10.1016/j.image.2004.06.003

VCRs/set-top-boxes and computers,and digital broadcast video delivered via terrestrial,cable or satellite networks,allowing digital TV and HDTV. While MPEG-1addressed coding of non-inter-laced video at lower resolutions and bit-rates[23] offering VHS-like video quality,MPEG-2ad-dressed coding of interlaced video at higher resolutions and bit-rates[24]enabling digital TV and HDTV with commensurate video quality.The MPEG-1standard was completed in1992while the MPEG-2standard was completed in1994.At the time of their completion they represented a timely as well as practical,state-of-the-art techni-cal solution[3,8,15,19,23,24,27,33,34],consistent with the cost/performance tradeoffs of the pro-ducts intended within the context of implementa-tion technology available.The actual impact of these standards in terms of inexpensive consumer products and market penetration took at least 5years from the time of completion of these standards.

MPEG-4was launched to address a new generation of multimedia applications and ser-vices.The core of the MPEG-4standard[29,11] was developed during a5-year period(1995–1999), however MPEG-4is a living standard with new parts added continuously as and when technology exists to address evolving applications.The pre-mise behind MPEG-4was future interactive multimedia applications and services such as interactive TV,Internet video etc where access to coded audio and video objects might be needed. The MPEG-4standard consists of many more parts besides the traditional,audio,video systems and conformance parts of earlier standards.Our discussion in this paper is however limited only to video.The MPEG-4video standard is designed as a toolkit standard with the capability to allow coding and thus access to individual objects, scalability of coded objects,transmission of coded video objects on error prone networks,as well as ef?cient coding of video objects.Further,MPEG-4 video also allows higher ef?ciency coding(than MPEG-1and MPEG-2)of rectangular video without the necessity of dividing a scene into video objects prior to coding.The signi?cant advances in core video standard referred to as MPEG-4part2,were achieved in the capability of coding of video objects while at the same time it clearly did improve coding ef?ciency over earlier standards.From coding ef?ciency standpoint, MPEG-4video was evolutionary in nature as it built on coding structure of MPEG-2and H.263 standards and adding enhanced/new tools but within the same coding structure.Thus,MPEG-4 part2offers a modest coding gain but only at the expense of a modest increase in complexity.The expectation was that since object-based video was the main focus,increase in complexity could be only justi?ed for those applications only,not for pure rectangular video applications.

In the meantime,while highly interactive multi-media applications appear farther into the future than anticipated,there seems to be an inexhaus-tible demand for much higher compression to enable with as best video quality as possible, practical applications such as internet multimedia, wireless video,personal video recorders,video-on-demand,and videoconferencing.The H.264/ MPEG-4AVC standard[14]is a new state-of-the-art video coding standard that addresses aforementioned applications.The core of this standard was completed in the form of?nal draft international standard(FDIS)in June2003while an extension for professional applications is currently in progress.It promises signi?cantly higher compression than earlier standards.The standard evolved from the original very promising work performed by ITU-T VCEG in their H.26L project over the period of1999–2001,and with MPEG joining the effort in late2001,a joint team of ITU-T VCEG and ISO MPEG experts was established for co-developing the standard.The resulting joint standard is called H.264by VCEG and MPEG-4part10by MPEG(as mentioned earlier,the original MPEG-4video is referred to as MPEG-4part2).Another name for this standard is MPEG-4advanced video coding(AVC)stan-dard.Further,informally it is also referred to as Joint Video Team(JVT)standard as it resulted from collaborative effort of the VCEG and MPEG standards committees.Regardless of the preferred name,the standard achieves clearly higher compression ef?ciency,often quoted as,up to a factor of2[38],over the MPEG-2video standard. As one would expect,the increase in compression

A.Puri et al./Signal Processing:Image Communication19(2004)793–849 794

ef?ciency comes at the cost of substantial increase in complexity,often quoted as factor of4for the decoder,while encoding complexity may be as high as factor of9over MPEG-2.Further,the exact improvement as well as the resulting com-plexity depends on the pro?le(subset)of the standard implemented,the choice of which is application dependent.It is worth noting that the standard uses the familiar motion compensated coding structure of earlier standards,a number of re?nements to existing tools in earlier standards, as well as key new tools and coding optimization. Also,the coding performance bene?ts of indivi-dual tools are much more scene and bit-rate dependent,and different tools differ signi?cantly in performance/complexity tradeoffs they offer. While the standard is discussed in a number of excellent papers[39,17],in this paper we address issues of practical video coding in an effort to help make informed selection of coding tools/pro?les and parameters in coder design.The rest of the paper covers the various aspects as follows.

Section2of this paper presents an overview of the H.264/MPEG-4AVC standard.It presents the coding structure,lists too …… 此处隐藏：48451字，全部文档内容请下载后查看。喜欢就下载吧 ……