An optimization scheme for quadtree-structured prediction and residual encoding in HEVC

Guifen Tian, Satoshi Goto

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

In the High Efficiency Video Coding (HEVC), block sizes for hybrid-prediction and residual encoding are recursively selected by using a quadtree structure. This technique requires enormous computational complexity. Nevertheless, the exhaustive quadtree-based partition is not always necessary. This paper fully utilizes all-zero residual blocks to accelerate HEVC encoding process. A near-sufficient condition is derived to detect variable-sized all-zero blocks. For these blocks, DCT and quantization can be skipped. Moreover, a novel PU pruning technique based on allzero block is presented to constrain prediction units (PU) which have little contribution to RD performance. Experiments on a wide range of videos show that proposed scheme can reduce up to 73.42% and an average of 53.37% computational complexity for HEVC encoder with only trivial loss in PSNR and rate.

Original languageEnglish
Title of host publicationIEEE Asia-Pacific Conference on Circuits and Systems, Proceedings, APCCAS
Pages547-550
Number of pages4
DOIs
Publication statusPublished - 2012
Event2012 IEEE Asia Pacific Conference on Circuits and Systems, APCCAS 2012 - Kaohsiung
Duration: 2012 Dec 22012 Dec 5

Other

Other2012 IEEE Asia Pacific Conference on Circuits and Systems, APCCAS 2012
CityKaohsiung
Period12/12/212/12/5

Fingerprint

Image coding
Computational complexity
Experiments

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Tian, G., & Goto, S. (2012). An optimization scheme for quadtree-structured prediction and residual encoding in HEVC. In IEEE Asia-Pacific Conference on Circuits and Systems, Proceedings, APCCAS (pp. 547-550). [6419093] https://doi.org/10.1109/APCCAS.2012.6419093

An optimization scheme for quadtree-structured prediction and residual encoding in HEVC. / Tian, Guifen; Goto, Satoshi.

IEEE Asia-Pacific Conference on Circuits and Systems, Proceedings, APCCAS. 2012. p. 547-550 6419093.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Tian, G & Goto, S 2012, An optimization scheme for quadtree-structured prediction and residual encoding in HEVC. in IEEE Asia-Pacific Conference on Circuits and Systems, Proceedings, APCCAS., 6419093, pp. 547-550, 2012 IEEE Asia Pacific Conference on Circuits and Systems, APCCAS 2012, Kaohsiung, 12/12/2. https://doi.org/10.1109/APCCAS.2012.6419093
Tian G, Goto S. An optimization scheme for quadtree-structured prediction and residual encoding in HEVC. In IEEE Asia-Pacific Conference on Circuits and Systems, Proceedings, APCCAS. 2012. p. 547-550. 6419093 https://doi.org/10.1109/APCCAS.2012.6419093
Tian, Guifen ; Goto, Satoshi. / An optimization scheme for quadtree-structured prediction and residual encoding in HEVC. IEEE Asia-Pacific Conference on Circuits and Systems, Proceedings, APCCAS. 2012. pp. 547-550
@inproceedings{7b2841339be94ff69008da29b8bbde25,
title = "An optimization scheme for quadtree-structured prediction and residual encoding in HEVC",
abstract = "In the High Efficiency Video Coding (HEVC), block sizes for hybrid-prediction and residual encoding are recursively selected by using a quadtree structure. This technique requires enormous computational complexity. Nevertheless, the exhaustive quadtree-based partition is not always necessary. This paper fully utilizes all-zero residual blocks to accelerate HEVC encoding process. A near-sufficient condition is derived to detect variable-sized all-zero blocks. For these blocks, DCT and quantization can be skipped. Moreover, a novel PU pruning technique based on allzero block is presented to constrain prediction units (PU) which have little contribution to RD performance. Experiments on a wide range of videos show that proposed scheme can reduce up to 73.42{\%} and an average of 53.37{\%} computational complexity for HEVC encoder with only trivial loss in PSNR and rate.",
author = "Guifen Tian and Satoshi Goto",
year = "2012",
doi = "10.1109/APCCAS.2012.6419093",
language = "English",
isbn = "9781457717291",
pages = "547--550",
booktitle = "IEEE Asia-Pacific Conference on Circuits and Systems, Proceedings, APCCAS",

}

TY - GEN

T1 - An optimization scheme for quadtree-structured prediction and residual encoding in HEVC

AU - Tian, Guifen

AU - Goto, Satoshi

PY - 2012

Y1 - 2012

N2 - In the High Efficiency Video Coding (HEVC), block sizes for hybrid-prediction and residual encoding are recursively selected by using a quadtree structure. This technique requires enormous computational complexity. Nevertheless, the exhaustive quadtree-based partition is not always necessary. This paper fully utilizes all-zero residual blocks to accelerate HEVC encoding process. A near-sufficient condition is derived to detect variable-sized all-zero blocks. For these blocks, DCT and quantization can be skipped. Moreover, a novel PU pruning technique based on allzero block is presented to constrain prediction units (PU) which have little contribution to RD performance. Experiments on a wide range of videos show that proposed scheme can reduce up to 73.42% and an average of 53.37% computational complexity for HEVC encoder with only trivial loss in PSNR and rate.

AB - In the High Efficiency Video Coding (HEVC), block sizes for hybrid-prediction and residual encoding are recursively selected by using a quadtree structure. This technique requires enormous computational complexity. Nevertheless, the exhaustive quadtree-based partition is not always necessary. This paper fully utilizes all-zero residual blocks to accelerate HEVC encoding process. A near-sufficient condition is derived to detect variable-sized all-zero blocks. For these blocks, DCT and quantization can be skipped. Moreover, a novel PU pruning technique based on allzero block is presented to constrain prediction units (PU) which have little contribution to RD performance. Experiments on a wide range of videos show that proposed scheme can reduce up to 73.42% and an average of 53.37% computational complexity for HEVC encoder with only trivial loss in PSNR and rate.

UR - http://www.scopus.com/inward/record.url?scp=84874163726&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84874163726&partnerID=8YFLogxK

U2 - 10.1109/APCCAS.2012.6419093

DO - 10.1109/APCCAS.2012.6419093

M3 - Conference contribution

AN - SCOPUS:84874163726

SN - 9781457717291

SP - 547

EP - 550

BT - IEEE Asia-Pacific Conference on Circuits and Systems, Proceedings, APCCAS

ER -