Use of motion estimation to improve video watermarking for MPEG encoders

Isao Echizen, Hiroshi Yoshiura, Yasuhiro Fujii, Satoru Tezuka

Research output: Chapter in Book/Report/Conference proceedingChapter

8 Citations (Scopus)

Abstract

Maintaining picture quality is an essential requirement for video watermarking, and this paper proposes a new method that uses motion estimation (block-matching techniques) to preferentially allocate watermarks to picture contents with motion based on our criterion for estimating watermark imperceptibility that takes motion into account. Experimental evaluations have shown that the new method significantly reduces degradation in picture quality and that, for the same picture quality after MPEG encoding and decoding, 30% more watermarks can be embedded than with the previous method. This new method would be most effective when implemented in MPEG encoders, which already have the necessary motion estimation functions.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
EditorsTon Kalker, Ingemar J. Cox, Yong Man Ro
PublisherSpringer Verlag
Pages184-199
Number of pages16
ISBN (Print)354021061X
DOIs
Publication statusPublished - 2004
Externally publishedYes

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2939
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Use of motion estimation to improve video watermarking for MPEG encoders'. Together they form a unique fingerprint.

Cite this