Hand Written English Character Recognition using Column-wise Segmentation of Image Matrix (CSIM)
Rakesh Kumar Mandal, Nilotpal Manna, West Bengal · 2012
Research is going on to develop both hardware and software to recognize handwritten characters easily and accurately. Artificial Neural Network (ANN) is a very efficient method for recognizing handwritten characters. Attempts have already been made to recognize English alphabets using similar type of methods. A new method has been tried, in this paper, to improve the performance of the previously applied methods. The input image matrix is compressed into a lower dimension matrix in order to reduce non significant elements of the image matrix. The compressed matrix is segmented column-wise. Each column of a particular image matrix is mapped to identical patterns for recognizing a particular character. Majority of a known pattern decides the existence of a particular character.