Title :
An improved reduced candidate modes approach for HEVC intra prediction
Author :
Kanayah Saurty;Pierre Clarel Catherine;K. M. S. Soyjaudah
Author_Institution :
Faculty of Information and Communication Technologies, Universite des Mascareignes, Pamplemousses, Mauritius
Abstract :
High Efficiency Video Coding (HEVC) is currently the latest standard in video compression. HEVC aims at compression of video at half the bit rate with the same quality or at the same bit rate with an improved quality compared to its predecessor, H.264. However, the increase in prediction modes (35) in HEVC has significantly affected the encoder complexity. In this paper, a reduced number of candidate modes is used in the Rate Distortion Optimization (RDO) process using a threshold which is based on the difference between the cost of the first mode and that of the last selected mode in the rough mode decision process. To the best of knowledge of the authors, the simulation results based on the proposed method outperform those from all other works that also make use of reduced modes. The method indeed yields an average of 41.8% time savings achievement along with an acceptable loss in quality (0.077 dB) and a negligible increase in bit rate (0.07%) compared to HM 10.
Keywords :
"Encoding","Bit rate","Complexity theory","Video coding","Intelligent systems","Rate-distortion","Transforms"
Conference_Titel :
SAI Intelligent Systems Conference (IntelliSys), 2015
DOI :
10.1109/IntelliSys.2015.7361225