Contents-aware partitioning algorithm for parallel high efficiency video coding

Kyungah Kim, Won Woo Ro

Research output: Contribution to journalArticle

Abstract

We introduce a new parallelization method for high-efficiency video coding (HEVC), which resolves the shortcomings of the existing tile-based parallel processing method. The parallel HEVC performs encoding by dividing a frame into numerous parallel units. This decreases the compression efficiency compared with sequential HEVC, because it artificially breaks the data correlation within a frame, which is called the parallelization overhead. The traditional parallel techniques such as Tiles and wavefront parallel processing (WPP) inherently introduce a high parallelization overhead because they simply divide a frame statically without considering the contents of the frame. The proposed new parallel encoding scheme resolves such problems by partitioning a frame based on the meaningful contents. In order to analyze the correlations within a frame and define the contents, the features within a frame are first extracted and clustered. In the feature clustering algorithm, two factors are considered to balance the workload between parallel units: (1) the number of features in each cluster and (2) the number of coding tree units (CTU) occupied by each cluster. The frame is partitioned based on the result of clustering, and the partitions are encoded in parallel. The proposed scheme achieves a bit-saving of up to 7.21%, with an average of 3.71%, along with an average time-saving of 20.50% compared to the Tiles technique.

Original languageEnglish
Pages (from-to)11427-11442
Number of pages16
JournalMultimedia Tools and Applications
Volume78
Issue number9
DOIs
Publication statusPublished - 2019 May 1

Fingerprint

Image coding
Tile
Wavefronts
Processing
Clustering algorithms

All Science Journal Classification (ASJC) codes

  • Software
  • Media Technology
  • Hardware and Architecture
  • Computer Networks and Communications

Cite this

@article{6dff2d069abf42dfa518cbec961997f0,
title = "Contents-aware partitioning algorithm for parallel high efficiency video coding",
abstract = "We introduce a new parallelization method for high-efficiency video coding (HEVC), which resolves the shortcomings of the existing tile-based parallel processing method. The parallel HEVC performs encoding by dividing a frame into numerous parallel units. This decreases the compression efficiency compared with sequential HEVC, because it artificially breaks the data correlation within a frame, which is called the parallelization overhead. The traditional parallel techniques such as Tiles and wavefront parallel processing (WPP) inherently introduce a high parallelization overhead because they simply divide a frame statically without considering the contents of the frame. The proposed new parallel encoding scheme resolves such problems by partitioning a frame based on the meaningful contents. In order to analyze the correlations within a frame and define the contents, the features within a frame are first extracted and clustered. In the feature clustering algorithm, two factors are considered to balance the workload between parallel units: (1) the number of features in each cluster and (2) the number of coding tree units (CTU) occupied by each cluster. The frame is partitioned based on the result of clustering, and the partitions are encoded in parallel. The proposed scheme achieves a bit-saving of up to 7.21{\%}, with an average of 3.71{\%}, along with an average time-saving of 20.50{\%} compared to the Tiles technique.",
author = "Kyungah Kim and Ro, {Won Woo}",
year = "2019",
month = "5",
day = "1",
doi = "10.1007/s11042-018-6619-8",
language = "English",
volume = "78",
pages = "11427--11442",
journal = "Multimedia Tools and Applications",
issn = "1380-7501",
publisher = "Springer Netherlands",
number = "9",

}

Contents-aware partitioning algorithm for parallel high efficiency video coding. / Kim, Kyungah; Ro, Won Woo.

In: Multimedia Tools and Applications, Vol. 78, No. 9, 01.05.2019, p. 11427-11442.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Contents-aware partitioning algorithm for parallel high efficiency video coding

AU - Kim, Kyungah

AU - Ro, Won Woo

PY - 2019/5/1

Y1 - 2019/5/1

N2 - We introduce a new parallelization method for high-efficiency video coding (HEVC), which resolves the shortcomings of the existing tile-based parallel processing method. The parallel HEVC performs encoding by dividing a frame into numerous parallel units. This decreases the compression efficiency compared with sequential HEVC, because it artificially breaks the data correlation within a frame, which is called the parallelization overhead. The traditional parallel techniques such as Tiles and wavefront parallel processing (WPP) inherently introduce a high parallelization overhead because they simply divide a frame statically without considering the contents of the frame. The proposed new parallel encoding scheme resolves such problems by partitioning a frame based on the meaningful contents. In order to analyze the correlations within a frame and define the contents, the features within a frame are first extracted and clustered. In the feature clustering algorithm, two factors are considered to balance the workload between parallel units: (1) the number of features in each cluster and (2) the number of coding tree units (CTU) occupied by each cluster. The frame is partitioned based on the result of clustering, and the partitions are encoded in parallel. The proposed scheme achieves a bit-saving of up to 7.21%, with an average of 3.71%, along with an average time-saving of 20.50% compared to the Tiles technique.

AB - We introduce a new parallelization method for high-efficiency video coding (HEVC), which resolves the shortcomings of the existing tile-based parallel processing method. The parallel HEVC performs encoding by dividing a frame into numerous parallel units. This decreases the compression efficiency compared with sequential HEVC, because it artificially breaks the data correlation within a frame, which is called the parallelization overhead. The traditional parallel techniques such as Tiles and wavefront parallel processing (WPP) inherently introduce a high parallelization overhead because they simply divide a frame statically without considering the contents of the frame. The proposed new parallel encoding scheme resolves such problems by partitioning a frame based on the meaningful contents. In order to analyze the correlations within a frame and define the contents, the features within a frame are first extracted and clustered. In the feature clustering algorithm, two factors are considered to balance the workload between parallel units: (1) the number of features in each cluster and (2) the number of coding tree units (CTU) occupied by each cluster. The frame is partitioned based on the result of clustering, and the partitions are encoded in parallel. The proposed scheme achieves a bit-saving of up to 7.21%, with an average of 3.71%, along with an average time-saving of 20.50% compared to the Tiles technique.

UR - http://www.scopus.com/inward/record.url?scp=85053881174&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85053881174&partnerID=8YFLogxK

U2 - 10.1007/s11042-018-6619-8

DO - 10.1007/s11042-018-6619-8

M3 - Article

VL - 78

SP - 11427

EP - 11442

JO - Multimedia Tools and Applications

JF - Multimedia Tools and Applications

SN - 1380-7501

IS - 9

ER -