TY - GEN
T1 - Fast Multi-Type Tree Partition for H.266/VVC Inter Coding
AU - Ciou, Yi Sheng
AU - Chen, Mei Juan
AU - Yeh, Chia Hung
AU - Lu, Chen Rung
AU - Hsieh, Meng Chun
AU - Lo, Chen
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Versatile Video Coding (H.266/VVC), the latest international video coding standard, effectively encodes ultra-high-definition video. Compared to the quadtree (QT) split of High Efficiency Video Coding (H.265/HEVC), the quadtree with nested multi-type tree (QT-MTT) structure applying binary tree (BT) and ternary tree (TT) coding blocks in H.266/VVC provides various coding unit (CU) partitions. The functionality improves the compression performance but greatly increases encoding complexity. Therefore, the acceleration of encoding is an essential issue. In this paper, the image feature extracted by edge detection is utilized to make fast decisions about the TT split and direction in the MTT structure of CU. Moreover, the maximum partition depth of the coding tree unit is predicted from the reference frames for the early termination of the encoding process. Experimental results demonstrate that the coding time of the proposed algorithm is reduced while the bitrate is well-maintained under random access configuration. The proposed method significantly outperforms the previous work.
AB - Versatile Video Coding (H.266/VVC), the latest international video coding standard, effectively encodes ultra-high-definition video. Compared to the quadtree (QT) split of High Efficiency Video Coding (H.265/HEVC), the quadtree with nested multi-type tree (QT-MTT) structure applying binary tree (BT) and ternary tree (TT) coding blocks in H.266/VVC provides various coding unit (CU) partitions. The functionality improves the compression performance but greatly increases encoding complexity. Therefore, the acceleration of encoding is an essential issue. In this paper, the image feature extracted by edge detection is utilized to make fast decisions about the TT split and direction in the MTT structure of CU. Moreover, the maximum partition depth of the coding tree unit is predicted from the reference frames for the early termination of the encoding process. Experimental results demonstrate that the coding time of the proposed algorithm is reduced while the bitrate is well-maintained under random access configuration. The proposed method significantly outperforms the previous work.
UR - http://www.scopus.com/inward/record.url?scp=85138737311&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85138737311&partnerID=8YFLogxK
U2 - 10.1109/ICCE-Taiwan55306.2022.9869026
DO - 10.1109/ICCE-Taiwan55306.2022.9869026
M3 - Conference contribution
AN - SCOPUS:85138737311
T3 - Proceedings - 2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022
SP - 471
EP - 472
BT - Proceedings - 2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2022 IEEE International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2022
Y2 - 6 July 2022 through 8 July 2022
ER -