Text detection based on wavelet transform

Jie He*, Fei Gao, Qing Nie

*Corresponding author for this work

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

Abstract

In this paper, an algorithm is proposed which can automatically detect text embedded in images and video frames in different sizes, fonts and colors.It is performed by these steps: Firstly, it performed wavelet transform in different color channels; Secondly, a sliding window is moved over the transformed images, estimate feature vector of each window; then the k-means algorithm is chosen to classify the image into text candidates and background. Finally, the detected text candidates undergo the connected-component-based analysis to refine their location.

Original languageEnglish
Title of host publicationChina-Ireland International Conference on Information and Communications Technologies, CIICT 2008
Pages82-84
Number of pages3
Edition544 CP
DOIs
Publication statusPublished - 2008
EventChina-Ireland International Conference on Information and Communications Technologies, CIICT 2008 - Beijing, China
Duration: 26 Sept 200828 Sept 2008

Publication series

NameIET Conference Publications
Number544 CP

Conference

ConferenceChina-Ireland International Conference on Information and Communications Technologies, CIICT 2008
Country/TerritoryChina
CityBeijing
Period26/09/0828/09/08

Keywords

  • Feature vector
  • K-means
  • Text detecting
  • Wavelet transform

Fingerprint

Dive into the research topics of 'Text detection based on wavelet transform'. Together they form a unique fingerprint.

Cite this