A new video images text localization approach based on a fast hough transform

  • Authors:

  • Bassem Bouaziz;Walid Mahdi;Abdelmajid Ben Hamadou

  • Affiliations:

  • Higher Institute of Computer Science and Multimedia, MIRACL, Multimedia InfoRmation system and Advanced Computing Laboratory, Sfax, Tunisia;Higher Institute of Computer Science and Multimedia, MIRACL, Multimedia InfoRmation system and Advanced Computing Laboratory, Sfax, Tunisia;Higher Institute of Computer Science and Multimedia, MIRACL, Multimedia InfoRmation system and Advanced Computing Laboratory, Sfax, Tunisia

  • Venue:
  • ICIAR'06 Proceedings of the Third international conference on Image Analysis and Recognition - Volume Part I
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

The amount of digital video data is increasing over the world. It highlights the need for efficient algorithms that can retrieve this data by content. The full use of this media is currently limited by the opaque nature of the video which prevents content-based access. To facilitate video indexing and browsing, it is essential to allow non-linear access, especially for long programs. This can be achieved by identifying semantic description captured automatically from video story structure. Among these descriptions, text within video frames is considered as rich features that enable a good way for browsing video. In this paper we propose a fast Hough transformation based approach for automatic video frames text localization. Experimental results, that we drove on a large sample of video images issued from portions of news broadcasts, sports program, advertisements and movies, shows that our method is very efficient, capable to locate text regions with different character sizes, directions and styles even in case of complex image background.