Fast Coding Unit Size Decision Based on Probabilistic Graphical Model in High Efficiency Video Coding Inter Prediction

Xiantao JIANG  Tian SONG  Wen SHI  Takafumi KATAYAMA  Takashi SHIMAMOTO  Lisheng WANG  

IEICE TRANSACTIONS on Information and Systems   Vol.E99-D   No.11   pp.2836-2839
Publication Date: 2016/11/01
Publicized: 2016/08/08
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2015EDL8237
Type of Manuscript: LETTER
Category: Image Processing and Video Processing
HEVC,  CU size,  encoding complexity,  probability graphical model,  

Full Text: PDF(220.2KB)>>
Buy this Article

In this work, a high efficiency coding unit (CU) size decision algorithm is proposed for high efficiency video coding (HEVC) inter coding. The CU splitting or non-splitting is modeled as a binary classification problem based on probability graphical model (PGM). This method incorporates two sub-methods: CU size termination decision and CU size skip decision. This method focuses on the trade-off between encoding efficiency and encoding complexity, and it has a good performance. Particularly in the high resolution application, simulation results demonstrate that the proposed algorithm can reduce encoding time by 53.62%-57.54%, while the increased BD-rate are only 1.27%-1.65%, compared to the HEVC software model.