Knowledge-supported Segmentation and Semantic Contents Extraction from MPEG Videos for Highlight-based Annotation, Indexing and Retrieval

Jinchang Ren, Juan Chen, Jianmin Jiang , Stan S. Ipson

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Automatic recognition of highlights from videos is a fundamental and challenging problem for content-based indexing and retrieval applications. In this paper, we propose techniques to solve this problem by using knowledge supported extraction of semantic contents, and compressed-domain processing is employed for efficiency. Firstly, video shots are detected by using knowledge-supported rules. Then, human objects are detected via statistical skindetection. Meanwhile, camera motion like zoom in is identified. Finally, highlights of zooming in human objects are extracted and used for annotation, indexing and retrieval of the whole videos. Results from large data of test videos have demonstrated the accuracy and robustness of the proposed techniques.
Original languageEnglish
Title of host publicationAdvanced intelligent computing theories and applications
Subtitle of host publicationwith aspects of theoretical and methodological issues
EditorsDe-Shuang Huang, Donald C. Wunsch II, Daniel S. Levine, Kang-Hyun Jo
Place of PublicationBerlin
PublisherSpringer
Pages258-265
Number of pages8
ISBN (Print)9783540874409
DOIs
Publication statusPublished - 2008
Event4th International Conference on Intelligent Computing, ICIC 2008 - Shanghai, China
Duration: 15 Sept 200818 Sept 2008

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume5226
ISSN (Print)0302-9743

Conference

Conference4th International Conference on Intelligent Computing, ICIC 2008
Country/TerritoryChina
CityShanghai
Period15/09/0818/09/08

Keywords

  • content-based indexing
  • video segmentation
  • semantic
  • highlights extraction
  • skin detection
  • camera motion
  • MPEG
  • shot detection

Fingerprint

Dive into the research topics of 'Knowledge-supported Segmentation and Semantic Contents Extraction from MPEG Videos for Highlight-based Annotation, Indexing and Retrieval'. Together they form a unique fingerprint.

Cite this