Handwritten Digit Recognition Using K-Nearest Neighbour Classifier

This paper presents a new approach to off-line handwritten digit recognition based on structural features which is not required thinning operation and size normalization technique. In this paper uses four different types of structural features namely, number of holes, water reservoirs in four directions, maximum profile distances in four directions, and fill-hole density for the recognition of digits. The digit recognition system mainly depends on which kinds of features are used. The main objective of this paper is to provide efficient and reliable techniques for recognition of handwritten digits. A Euclidean minimum distance criterion is used to find minimum distances and k-nearest neighbor classifier is used to classify the digits. A MNIST database is used for both training and testing the system. 5000 images are used to test the proposed method a total 5000 numeral images are tested and got 96.94% recognition rate.