Classification of Handwritten Marathi Numbers Using Euclidean Distance

In this paper we propose a system for recognition of Marathi handwritten numbers. In this system we develop an algorithm in which we employed Euclidean distance technique for recognition. This algorithm tested on our own database in which we collected 0-9 handwritten Marathi numbers and this handwritten numbers converted in digitization using scanner of brand UMAX model Astra 5600. The handwritten font size of each numbers for each person is different and we made same size, by comparing query/test image of number with database image of numbers by removing rows and columns from bigger size of row and column. Also we compare two techniques of removing rows and columns from centre and removed rows and columns from outer sides (i.e. from top, bottom, left and right side) of database images of numbers for calculation of Euclidean distance. The algorithm tested in Matlab tool and the result is shown in table.