An efficient method for text detection in video based on stroke width similarity

Viet Cuong Dinh, Soo Chun Seong, Cha Seungwook, Ryu Hanjin, Sanghoon Sull

Research output: Chapter in Book/Report/Conference proceedingConference contribution

20 Citations (Scopus)


Text appearing in video provides semantic knowledge and significant information for video indexing and retrieval system. This paper proposes an effective method for text detection in video based on the similarity in stroke width of text (which is defined as the distance between two edges of a stroke). From the observation that text regions can be characterized by a dominant fixed stroke width, edge detection with local adaptive thresholds is first devised to keep text- while reducing background-regions. Second, morphological dilation operator with adaptive structuring element size determined by stroke width value is exploited to roughly localize text regions. Finally, to reduce false alarm and refine text location, a new multi-frame refinement method is applied. Experimental results show that the proposed method is not only robust to different levels of background complexity, but also effective to different fonts (size, color) and languages of text.

Original languageEnglish
Title of host publicationComputer Vision - ACCV 2007 - 8th Asian Conference on Computer Vision, Proceedings
PublisherSpringer Verlag
Number of pages10
EditionPART 1
ISBN (Print)9783540763857
Publication statusPublished - 2007
Event8th Asian Conference on Computer Vision, ACCV 2007 - Tokyo, Japan
Duration: 2007 Nov 182007 Nov 22

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume4843 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other8th Asian Conference on Computer Vision, ACCV 2007

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)


Dive into the research topics of 'An efficient method for text detection in video based on stroke width similarity'. Together they form a unique fingerprint.

Cite this