Handwritten devnagari character recognition using connected segments and minimum edit distance

Character recognition has long been a critical area of the OCR process. We introduce a new segmentation technique guided in part by the global characteristics of the handwriting. In this paper we have given a technique to identify Devnagari characters based on segmentation using various operators and converting image into a set of characters having definite prerequisite relationship. A method is developed for matching string with strings in database based on minimum edit distance string matching algorithm. This technique can be divided into four major steps. 1) Digitization of image. 2) Thinning 3) Segmentation of image 4) Coding the image based on connected segments. 5) Applying minimum edit distance string matching algorithm. Experimental results support the effectiveness of proposed idea.