Sample Adaptive Offset Parallelism in HEVC
We propose a parallelization method for SAO, in-loop filter of HEVC. SAO filtering proceeds along CTB lines and there exists data dependency between inside and outside of CTB boundaries. Data dependency makes data-level parallelization hard. In this paper
- PDF / 1,785,855 Bytes
- 7 Pages / 439.37 x 666.142 pts Page_size
- 105 Downloads / 186 Views
Abstract We propose a parallelization method for SAO, in-loop filter of HEVC. SAO filtering proceeds along CTB lines and there exists data dependency between inside and outside of CTB boundaries. Data dependency makes data-level parallelization hard. In this paper, we equally divided an entire frame into sub regions. With a little amount of memory, proposed method shows 1.9 times of performance enhancement in terms of processing time.
Keywords Sample adaptive offset SAO SAO parallelism HEVC parallelism Multi-core parallelism In-loop filter
1 Introduction Needs for the realistic video service, such as higher resolution and higher quality, are increasing with the developments of multimedia-related hardware and software technologies. In addition, requirements for the advanced video coding standard have arisen on the multimedia industrial market. Dependent on the demands, Joint E. Ryu J. Nam S. Lee H. Jo D. Sim (&) Computer Engineering, Kwangwoon University, Seoul, Republic of Korea e-mail: [email protected] E. Ryu e-mail: [email protected] J. Nam e-mail: [email protected] S. Lee e-mail: [email protected] H. Jo e-mail: [email protected]
J. J. (Jong Hyuk) Park et al. (eds.), Multimedia and Ubiquitous Engineering, Lecture Notes in Electrical Engineering 240, DOI: 10.1007/978-94-007-6738-6_137, Ó Springer Science+Business Media Dordrecht(Outside the USA) 2013
1113
1114
E. Ryu et al.
Collaborative Team on Video Coding (JCT-VC) which is composed by ISO/IEC Moving Picture Experts Group (MPEG) and ITU-T Video Coding Experts Group (VCEG) has been standardizing HEVC, whose target coding efficiency is twice better than that of H.264/AVC [1]. To introduce the state-of-art video codec to various multimedia industries as soon as standardization process of HEVC is completed, researches on algorithms lowering computational complexity and methods of optimization for real time processing are indispensable. Even during standardizing process, parallel processing schemes for real-time processing have been contributed and accepted for HEVC. Thinking of trends on multi-core and various mobile devices and high complexity of HEVC, researches on parallelization for HEVC should be done. Lowering quantization error of reconstructed frames, SAO has dependency in pixel decoding process. To develop SAO with parallel scheme, we should think of a feature of SAO dependency. In this paper, we propose a parallel core assigning method which resolves CTBlevel dependency of SAO. We assign the same number of CTBs into each core. This paper is organized as follows. In Sect. 2, SAO method is introduced. In Sect. 3, dependency between adjacent CTBs is analyzed, and the proposed method is shown. Experimental results and evaluation are given in Sect. 4. Finally, conclusions are stated in Sect. 5.
2 Sample Adaptive Offset for HEVC New in-loop filter, SAO has been adopted for HEVC. Different to the other tools of video codec, quantization process causes data loss between original and reconstructed videos. Larger transforms and longer-tap i
Data Loading...
