Persistent Link:
http://hdl.handle.net/10150/290586
Title:
Error concealment for packet video
Author:
Yu, Gong-san, 1962-
Issue Date:
1996
Publisher:
The University of Arizona.
Rights:
Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
Abstract:
Two new algorithms are proposed for error concealment in the presence of cell/packet loss for packet video. Under packet switching network environments like Ethernet or ATM, some video packets may be lost or discarded due to transmission delay or channel congestion. Because video is normally compressed, small data losses may completely destroy the received video signal. Thus, lost block recovery or concealment becomes an essential issue for video communication over networks. The proposed algorithms use block DCT coding with one pixel overlap. Without any assumption on the smoothness of block boundaries, the combination of these two algorithms provide consistently good performance in lost block recovery. Experimental results show that the quality of the recovered video is much better than that obtained with maximally smooth methods. One additional advantage of the one pixel overlap structure is that it removes blocking effects automatically. Blocking effects are annoying problems for all block transform based image or video CODECs. With simple averaging operations on the overlapped pixels, all blocking effects can be suppressed. One-pixel overlapped block coding generally decreases the compression ratio for a given quantization step size. However, increases in PSNR achieved by overlapped block coding largely compensate for this effect. Experimental results show that for a given PSNR, the decrease in compression ratio is negligible. To demonstrate the proposed algorithms, a standard CCITT H.261 software encoder and decoder were created using C++. With minor modification, the standard CODEC is easily combined with our proposed error concealment algorithms to provide high performance block loss recovery for packet video.
Type:
text; Dissertation-Reproduction (electronic)
Keywords:
Engineering, Electronics and Electrical.; Computer Science.
Degree Name:
Ph.D.
Degree Level:
doctoral
Degree Program:
Graduate College; Electrical and Computer Engineering
Degree Grantor:
University of Arizona
Advisor:
Marcellin, Michael W.

Full metadata record

DC FieldValue Language
dc.language.isoen_USen_US
dc.titleError concealment for packet videoen_US
dc.creatorYu, Gong-san, 1962-en_US
dc.contributor.authorYu, Gong-san, 1962-en_US
dc.date.issued1996en_US
dc.publisherThe University of Arizona.en_US
dc.rightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.en_US
dc.description.abstractTwo new algorithms are proposed for error concealment in the presence of cell/packet loss for packet video. Under packet switching network environments like Ethernet or ATM, some video packets may be lost or discarded due to transmission delay or channel congestion. Because video is normally compressed, small data losses may completely destroy the received video signal. Thus, lost block recovery or concealment becomes an essential issue for video communication over networks. The proposed algorithms use block DCT coding with one pixel overlap. Without any assumption on the smoothness of block boundaries, the combination of these two algorithms provide consistently good performance in lost block recovery. Experimental results show that the quality of the recovered video is much better than that obtained with maximally smooth methods. One additional advantage of the one pixel overlap structure is that it removes blocking effects automatically. Blocking effects are annoying problems for all block transform based image or video CODECs. With simple averaging operations on the overlapped pixels, all blocking effects can be suppressed. One-pixel overlapped block coding generally decreases the compression ratio for a given quantization step size. However, increases in PSNR achieved by overlapped block coding largely compensate for this effect. Experimental results show that for a given PSNR, the decrease in compression ratio is negligible. To demonstrate the proposed algorithms, a standard CCITT H.261 software encoder and decoder were created using C++. With minor modification, the standard CODEC is easily combined with our proposed error concealment algorithms to provide high performance block loss recovery for packet video.en_US
dc.typetexten_US
dc.typeDissertation-Reproduction (electronic)en_US
dc.subjectEngineering, Electronics and Electrical.en_US
dc.subjectComputer Science.en_US
thesis.degree.namePh.D.en_US
thesis.degree.leveldoctoralen_US
thesis.degree.disciplineGraduate Collegeen_US
thesis.degree.disciplineElectrical and Computer Engineeringen_US
thesis.degree.grantorUniversity of Arizonaen_US
dc.contributor.advisorMarcellin, Michael W.en_US
dc.identifier.proquest9706173en_US
dc.identifier.bibrecord.b34294077en_US
All Items in UA Campus Repository are protected by copyright, with all rights reserved, unless otherwise indicated.