Subscribe Now Subscribe Today
Science Alert
 
FOLLOW US:     Facebook     Twitter
Blue
   
Curve Top
Journal of Applied Sciences
  Year: 2013 | Volume: 13 | Issue: 9 | Page No.: 1537-1543
DOI: 10.3923/jas.2013.1537.1543
 
Facebook Twitter Digg Reddit Linkedin StumbleUpon E-mail
Vector Constraint and Ncc Based Chinese Document Image Mosaic
Lijing Tong, Quanyao Peng, Sam Li, Huiqun Zhao and Guoliang Zhan

Abstract:
Document images which captured by camera-based systems are often suffered from warping distortion that deteriorates the performance of current Optical Character Recognition (OCR) approaches. To overcome this difficulty and get better Chinese document image OCR recognition accuracy, a document image mosaic method based on vector constraint and Normalized Cross Correlation (NCC) is proposed. In this method, two images, one is reference image and the other is auxiliary image, are captured for the same warped document from left side and right side firstly. Then, for two reference points in the inflection point position in reference image, NCC registration method are used to find two matching points in the auxiliary image. During registration, the area limiting method and vector constraining method are proposed to improve the registration performance. At last, the two flatter parts of each image are mosaiced together. Experimental results show that the document image mosaic method based on vector constraining and NCC is more robust than the classical cross correlation registration method. The OCR recognition rate of the new mosaiced image is markedly higher than the two original distorted document images and the time cost is reduced.
PDF References Citation Report Citation
How to cite this article:

Lijing Tong, Quanyao Peng, Sam Li, Huiqun Zhao and Guoliang Zhan, 2013. Vector Constraint and Ncc Based Chinese Document Image Mosaic. Journal of Applied Sciences, 13: 1537-1543.

DOI: 10.3923/jas.2013.1537.1543

URL: https://scialert.net/abstract/?doi=jas.2013.1537.1543

 
COMMENT ON THIS PAPER
 
 
 

 

 
 
 
 
 
 
 
 
 

 
 
 
 
 
 
 

       

       

Curve Bottom