Use of artificial neural network in Bengali character recognition

This paper describes how the Bengali characters are processed, trained and then recognized with the use of a back propagation Artificial neural network. Recognition is done on isolated Bengali character. The size and the font used for the characters are similar in both training and classification of the network. The images are first converted into grayscale and then to binary images. These images are then scaled to a fit a pre-determined area with a fixed but significant number of pixels. By extracting the characteristics points we formed the feature vectors, which in this case is simply a series of 0s and 1s of fixed length. Finally, an Artificial Neural Network is chosen for the training and classification process. The steps are simple, and a network is chosen for the training and recognition process. Researchers involved in recognition of good quality printed text in different scripts around the world have reported drastic decrease in recognition accuracy due to presence of touching characters in the text. So recognition is done here with isolated printed characters with size independent.