Publication detail

Method for Real-Time Face Tracking in a Video Sequence

ŘÍHA, K. PŘINOSIL, J. FU, D. ZUKAL, M. KARÁSEK, J.

Original Title

Method for Real-Time Face Tracking in a Video Sequence

English Title

Method for Real-Time Face Tracking in a Video Sequence

Type

conference paper

Language

en

Original Abstract

The article presents a method for real-time tracking of face in video sequences. Algorithms for face detection are often very time-consuming and there are problems with their performance in real-time processing of video sequences or video streams. The algorithm presented in the article solves this problem using the optical flow technique. Face detection can be performed in a longer interval (in the order of hundreds of milliseconds) than the FPS of a given video sequence is performed. A special functionality for the evaluation of partial movements in the area of the face being tracked has been developed. The part of the article focused on testing the algorithm shows a high reduction of computation time with the high-accuracy determination of the face position preserved even in frames where the face detection algorithm was not applied.

English abstract

The article presents a method for real-time tracking of face in video sequences. Algorithms for face detection are often very time-consuming and there are problems with their performance in real-time processing of video sequences or video streams. The algorithm presented in the article solves this problem using the optical flow technique. Face detection can be performed in a longer interval (in the order of hundreds of milliseconds) than the FPS of a given video sequence is performed. A special functionality for the evaluation of partial movements in the area of the face being tracked has been developed. The part of the article focused on testing the algorithm shows a high reduction of computation time with the high-accuracy determination of the face position preserved even in frames where the face detection algorithm was not applied.

Keywords

Face Detection, Optical Flow, Tracking, Digital Video Sequence, Image Processing

RIV year

2012

Released

07.09.2012

Publisher

WSEAS Press

Location

Sliema, Malta

ISBN

978-1-61804-119-7

Book

Advances in Sensors, Signals, Visualisation, Imaging and Simulation

Edition

Recent Advances in Electrical Engineering Series

Edition number

6

Pages from

182

Pages to

187

Pages count

6

Documents

BibTex


@inproceedings{BUT93885,
  author="Kamil {Říha} and Jiří {Přinosil} and Dongmei {Fu} and Martin {Zukal} and Jan {Karásek}",
  title="Method for Real-Time Face Tracking in a Video Sequence",
  annote="The article presents a method for real-time tracking of face in video sequences. Algorithms for face detection are often very time-consuming and there are problems with their performance in real-time processing of video sequences or video streams. The algorithm presented in the article solves this problem using the optical flow technique. Face detection can be performed in a longer interval (in the order of hundreds of milliseconds) than the FPS of a given video sequence is performed. A special functionality for the evaluation of partial movements in the area of the face being tracked has been developed. The part of the article focused on testing the algorithm shows a high reduction of computation time with the high-accuracy determination of the face position preserved even in frames where the face detection algorithm was not applied.",
  address="WSEAS Press",
  booktitle="Advances in Sensors, Signals, Visualisation, Imaging and Simulation",
  chapter="93885",
  edition="Recent Advances in Electrical Engineering Series",
  howpublished="electronic, physical medium",
  institution="WSEAS Press",
  year="2012",
  month="september",
  pages="182--187",
  publisher="WSEAS Press",
  type="conference paper"
}