Fast Character Recognition Using Kohonen Neural Network

Abstract

This paper focuses at recognition of handwritten characters with the help of neural networks. Neural Network is designed and trained to classify and recognize handwritten characters. Various stages are involved in character recognition. Steps involves Image preprocessing, segmentation and feature extraction. Image acquisition is the first stage followed by noise filtering, smoothing and image normalization of scanned image. Segmentation divides the image into sub images and feature extraction elicit features from input image.