Video quality estimation for packet loss based on no-reference method

Junghyun Han, Yo Han Kim, Jangkeun Jeong, Jitae Shin

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

5 Citations (Scopus)

Abstract

Objective video-quality assessment can be divided into Full Reference (FR), Reduced Reference (RR), and No Reference (NR) methods. Among them NR quality assessment method does not require original video sequence, so it is suitable for the real-time video streaming service. This paper presents an NR-based estimation of mean square error (MSE) for video quality degradation due to various pattern packet losses in packet video network environment. Our proposed algorithm predicts initial error due to packet loss in each frame by using motion information of received data. Proposed method shows high Pearson correlation (more than 95%) to match actual MSE.

Original languageEnglish
Title of host publication12th International Conference on Advanced Communication Technology
Subtitle of host publicationICT for Green Growth and Sustainable Development, ICACT 2010 - Proceedings
Pages418-421
Number of pages4
Publication statusPublished - 2010
Externally publishedYes
Event12th International Conference on Advanced Communication Technology: ICT for Green Growth and Sustainable Development, ICACT 2010 - , Korea, Republic of
Duration: 2010 Feb 72010 Feb 10

Publication series

NameInternational Conference on Advanced Communication Technology, ICACT
Volume1
ISSN (Print)1738-9445

Other

Other12th International Conference on Advanced Communication Technology: ICT for Green Growth and Sustainable Development, ICACT 2010
Country/TerritoryKorea, Republic of
Period10/2/710/2/10

Keywords

  • Mean Square Error (MSE)
  • Motion vector
  • No-reference
  • Packet loss
  • Video quality estimation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Video quality estimation for packet loss based on no-reference method'. Together they form a unique fingerprint.

Cite this