A Radial Basis Function Neural Network to Recognize Handwritten Numerals with normalized moment features from skeletons

Handwritten numeral character recognition has been an intensive research in the field of artificial intelligence since many decades. This paper proposes a radial basis function neural network model for recognizing handwritten numerals. The geometric shape of handwritten numerals is described by computing a feature vector based on the skeleton of the images. The normalized central moment features are extracted from the skeleton of the images. Classification is performed with these normalized moment features by a radial basis function neural network. The novelty of this approach is that the normalized moment features from the skeletons gives good recognition rate than the contour images and thinned images with radial basis function neural network. The performance of the proposed work is computed from the error rate. Results of this proposed method on MNIST handwritten numeral database is reported.