Performance Evaluation of RBF, Cascade, Elman, Feed Forward and Pattern Recognition Network for Marathi Character Recognition with CLAHE Feature Extraction Method

Abstract

The purpose of this paper is to study, analyze and improve the performance of RBF, Cascade, Elman, Feed Forward and Pattern Recognition Networks using �Contrast-limited Adaptive Histogram Equalization method� of featureextraction. This work is divided in to two sections. In the earlier work, we have performed the performance analysis of RBF neural network, Cascade Neural network, Elman Neural network and Feed forward neural network for the character recognition of handwritten Marathi curve scripts using �Edge detection and Dilation method� of feature extraction. In this paper, we have applied the feature extraction methodknown as Contrast-limited Adaptive Histogram Equalization (CLAHE). This feature extraction method enhances the contrast of images by transforming the values in the intensity image. For this experiment, we have considered the six samples each of 48 Marathi characters. For every sampled character, the CLAHE feature extraction method is applied. Then we have studied and analyzed the performance of these five Neural Networks for character recognition. It is found that except Elman Network, the performance of rest of all the networks is increased.