TEXT DETECTION AND EXTRACTION FROM SCENE IMAGES USING WAVELET FEATURES

ANGADI S.A.¹*, KODABAGI M.M.²*, KAGAWADE V.C.³*, VINAYAKA HITTALAMANI⁴*
¹Department of Computer Science & Engineering, Basaveshwar Engineering College, Bagalkot, Karnataka, India
²Department of Computer Science & Engineering, Basaveshwar Engineering College, Bagalkot, Karnataka, India
³Department of Master of Computer Applications, Basaveshwar Engineering College, Bagalkot, Karnataka, India
⁴Department of Computer Science & Engineering, Basaveshwar Engineering College, Bagalkot, Karnataka, India
* Corresponding Author : vinayakahittalamani@gmail.com

Received : 12-04-2012     Accepted : 15-05-2012     Published : 21-05-2012
Volume : 4     Issue : 1       Pages : 409 - 409
Int J Mach Intell 4.1 (2012):409-409
DOI : http://dx.doi.org/10.9735/0975-2927.4.1.409-409

Keywords : Text detection, scene images, wavelet energy features, discriminant functions
Conflict of Interest : None declared

Text detection and extraction from scene images is useful in finding certain facts in supporting court of laws for solving crime problems and various forensic science applications. Text detection from scene images is a difficult and challenging problem due to various issues such as; varying font size, style, uneven illumination and other degradations. In this paper, a new method that uses wavelet energy features for text detection and extraction from scene images is presented. Initially, preprocessed binarized image will be divided into 50x50 blocks. Then, wavelet energy features are obtained from every block and potential text blocks are identified using newly defined discriminant functions. Further, the detected text blocks are merged and refined to extract text regions. The proposed method is robust and achieves a detection rate of 97% on a variety of scene images each of size 240x320.