An artificial neural network based segmentation algorithm for off-line handwriting recognition

Artificial Neural Networks (ANNs) have been successfully applied to Optical Character Recognition (OCR) yielding excellent results. This paper presents a method for segmentation of difficult handwriting with the use of conventional algorithms in conjunction with ANNs. The segmentation algorithm is heuristic in nature detecting important features which may represent a prospective segmentation point. An Artificial Neural Network is subsequently used to verify the authenticity of the segmentation points found by the algorithm. The C programming language, the SP2 supercomputer and a SUN workstation were used for the experiments. The algorithm has been tested on real-world handwriting obtained from the CEDAR database. Some preliminary experimental results are presented in this paper.