By Topic

Block adaptive binarisation of ill-conditioned business card images acquired in a PDA using a modified quadratic filter

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $31
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

3 Author(s)
Shin, K.T. ; Panel Design Team, LG Philips LCD Co. Ltd., Gumi ; Jang, I.H. ; Kim, N.C.

A block adaptive binarisation method using a modified quadratic filter (MQF) to binarise business card images having ill conditions of very poor resolution, weak and irregular illumination, shadow and noise acquired in hand-held personal digital assistant cameras, is proposed. In the proposedmethod, a business card image is first partitioned into blocks of 8times8, among which character blocks (CBs) are extracted based on their block activity in DCT domain. Each CB is next enhanced by the MQF that is a result of modifying the quadratic filter (QF) to be proper for locally adaptive enhancement. The enhanced CB is then binarised with the same threshold used in the enhancement so that the enhancement and binarisation are associated consistently with each other. All the remaining blocks are next classified as background blocks and binarised as background. A binarised image is finally obtained by tiling each binarised block at its original position. Experimental results show that the quality of binarised images obtained by the proposed method is much better than that by the conventional global binarisation (GB) using QF and that by Rodtook and Rangsanseri's (2001) pixel adaptive binarisation. It is also better than that by Sauvola and Pietikaumlinen's (2000) pixel adaptive binarisation and that by Yang and Yan's (2000) adaptive logical level binarisation. In addition, the proposed method yields about 32, 28.2, 5.6, and 6.4% improvement of character recognition rate over GB using QF (GB-QF), Rodtook and Rangsanseri's method, Sauvola and Pietikaumlinen's method, and Yang and Yan's method, respectively

Published in:

Image Processing, IET  (Volume:1 ,  Issue: 1 )