95 articles 

inria-00108401, version 1

Combination of Multiple Handwritten Text Line Recognition Systems with a Recursive Approach

Roman Bertolami () 1, Beat Halter () 1, Horst Bunke () 1

Tenth International Workshop on Frontiers in Handwriting Recognition (2006)

Abstract: In this paper we propose a novel method to combine the results of multiple text line recognition systems. The method uses a recursive approach and re-examines those parts in a text line which have been rejected based on the initial combination of the base recognisers' results. By means of the new method, the search space can be reduced, and therefore more accurate recognition results can be expected. Experiments conducted on the IAM database show that the proposed method is able to improve the recognition rate compared to a standard combination scheme.

  • 1:  Institute of Computer Science and Applied Mathematics (IAM)
  • Universität Bern
  • Domain : Computer Science/Document and Text Processing
    Computer Science/Computer Vision and Pattern Recognition
  • Keywords : Handwritten Text Line Recognition – Multiple Classifier Systems – Hidden Markov Models
  • Comment : http://www.suvisoft.com
 
  • inria-00108401, version 1
  • oai:hal.inria.fr:inria-00108401
  • From: 
  • Submitted on: Friday, 20 October 2006 16:12:12
  • Updated on: Friday, 20 October 2006 16:23:02