Offline automatic actor tracking in a movie

Chih Yang Lin, Hong Xia Xie, Shang Ming Wang, Pin Ming Su, Wen Thong Chang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Automatic extraction and tracking of actors in a movie presents a novel and challenging issue in the field of image and video processing. This paper presents a system that can efficiently track each actor in a movie and extract the scenes that feature a specific actor. The method's facial detection and recognition are based on Haar-like features and Eigenface for SVM, while the actor tracking proposes a novel approach using a B-A-P frame structure. We also modified the Camshift method to enhance tracking performance. Experimental results show that the proposed system can accurately detect and track actors and efficiently solve for the problem of object occlusion.

Original languageEnglish
Title of host publication2017 International Conference on Image and Vision Computing New Zealand, IVCNZ 2017
PublisherIEEE Computer Society
Pages1-5
Number of pages5
ISBN (Electronic)9781538642764
DOIs
StatePublished - 2 Jul 2017
Event2017 International Conference on Image and Vision Computing New Zealand, IVCNZ 2017 - Christchurch, New Zealand
Duration: 4 Dec 20176 Dec 2017

Publication series

NameInternational Conference Image and Vision Computing New Zealand
Volume2017-December
ISSN (Print)2151-2191
ISSN (Electronic)2151-2205

Conference

Conference2017 International Conference on Image and Vision Computing New Zealand, IVCNZ 2017
Country/TerritoryNew Zealand
CityChristchurch
Period4/12/176/12/17

Keywords

  • Object tracking
  • face detection
  • face recognition
  • scene change detection

Fingerprint

Dive into the research topics of 'Offline automatic actor tracking in a movie'. Together they form a unique fingerprint.

Cite this