
Fast prediction mode selection and CU partition for HEVC intra coding
Author(s) -
Fu Bin,
Zhang Qiangqing,
Hu Jianling
Publication year - 2020
Publication title -
iet image processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.401
H-Index - 45
eISSN - 1751-9667
pISSN - 1751-9659
DOI - 10.1049/iet-ipr.2019.0259
Subject(s) - computer science , encoder , reference software , coding (social sciences) , computational complexity theory , algorithm , algorithmic efficiency , coding tree unit , bit rate , context adaptive binary arithmetic coding , software , data compression , decoding methods , computer engineering , mathematics , statistics , programming language , operating system
Compared with the predecessor H.264/advanced video coding, high‐efficiency video coding (HEVC) is a new video coding standard with nearly double coding efficiency under the same coding quality. However, the computing complexity of HEVC increases sharply. To solve this problem, a fast algorithm for intra prediction mode selection based on mode grouping is proposed by reducing the number of modes entering rough mode decision. Moreover, the dual support vector machine is proposed to efficiently select the coding unit (CU) size, which is based on texture features of CU and sub‐CUs including content complexity and direction complexity. By using the new CU size selection algorithm, the encoder can confirm the split for complex CU and terminate the split for simple CU in advance, so as to reduce the computing complexity of CU size selection. The experimental results show that by employing the two fast algorithms in intra coding, it can save 42.80% encoding time, with 0.98% increment in bit rate and 0.018 dB loss of peak‐signal‐to‐noise ratio of luminance, compared with the reference software x265‐1.7.