An efficient H.264 based fine-granular-scalable video coding system Kemal Ugur, Panos Nasiopoulos, Rabab Ward Department of Electrical and Computer Engineering University of British Columbia, Vancouver, BC, CANADA {kemalu, panos, rababw @ece.ubc.ca}Authors Abstract—H.264, with its superior coding efficiency and network friendly design, has emerged as the newest international video standard and is expected to become the preferred codec for video broadcasting. Presently, there is an effort to add scalability to H.264, in order to offer a solution to network congestion and bandwidth variations. Proposals range from scalable subband extension to introduction of FGS to the H.264 standard. We propose a novel structure for the FGS layer that uses 4x4 Integer Transform, instead of 8x8 Discrete Cosine Transform (DCT), so that the same transform is used for both layers. We also propose a novel hierarchical algorithm to code macroblock header of FGS layer that uses less bits than the standard FGS algorithm and significantly increases the coding efficiency. When compared with the original FGS structure, the proposed structure uses 70% less bits for macroblock headers, has less complexity and has an increased PSNR of 0.7 dB on average. I. INTRODUCTION In networks used for video transmission environment, such as wireless networks and Internet, the available bandwidth for video transmission is not constant but varies over time [1]. Traditional video coding standards, whose objective is to optimize the quality of the video at a given bitrate, cannot cope with this bandwidth variation problem effectively. Scalable Video Coding techniques have been developed to more efficiently address this bandwidth variation problem. Recently, ISO/IEC Moving Picture Experts Group (MPEG) standardized Fine Granular Scalability (FGS) technology in the MPEG-4 Streaming Video Profile. The H.264 video coding standard includes a number of advances in video coding technology, making it highly efficient in terms of coding and network friendliness. H.264 standard achieves 50% bit rate savings when compared with previous standards [2, 3]. Recently, there is an effort to add scalability to H.264, to offer a solution to network congestion and bandwidth variations. Proposals range from scalable subband extension to intoroduction of FGS to the H.264 standard that include straightforward extension of FGS to H.264 [4, 5]. The latter proposals use the H.264 standard to code the base layer and the FGS standard, without any modifications, to code the enhancement layer. Such straightforward extension of FGS, which uses H.264 instead of MPEG-4 to code the base layer, is possible due to FGS’ design. However, it has serious drawbacks that significantly increase the complexity of the system due to the fundamentally different tools used in H.264 & MPEG-4. One drawback of a such implementation is the use of Discrete Cosine Transform (DCT) at the FGS enhancement layer of an H.264 based FGS system. Because H.264 standard replaced DCT with 4x4 Integer Transform, using DCT at the enhancement layer brings additional computational complexity to the H.264 based FGS system. This additional need for computing another transform significantly increases the complexity of both the encoder and decoder (particularly the latter). In this paper, we propose to use 4x4 Integer Transform at the FGS layer instead of 8x8 DCT. By using the same transform for both the FGS enhancement layer and H.264 base layers, we achieve lower complexity for both the encoder and decoder. However, the change in the type of transform, changes the FGS macroblock structure due to the difference in the transform sizes. The change in macroblock structure results in an increase in the number of bits for the Coded_Block_Pattern (CBP) that needs to be coded at every macroblock header, hence decreases the coding efficiency. To decrease the number of bits used for CBP, we propose a novel CBP coding algorithm, called Hierarchical CBP Coding. When used with our H.264 based FGS system, our proposed Hierarchical CBP Coding algorithm decreases the overhead by 70% on average, and it makes efficient usage of 4x4 Integer Transform at the enhancement layer possible. Experimental results show that, in addition to having less complexity, our proposed H.264 based FGS system achieves a PSNR gaing of 0.7 dB on average. This paper is organized as follows: Section 2 briefly describes the MPEG-4 FGS. Section 3 presents our H.264 based FGS system and our novel Hierarchical CBP Coding Algorithm. A complexity analysis of the proposed scheme is also presented in Section 3. Section 4 presents the experimental results and conclusions are given in Section 5. 0-7803-9333-3/05/$20.00 ©2005 IEEE SIPS 2005 399