Signature recognition using Krawtchouk moments

An Efficient signature recognition system is proposed in this paper. The input signature image is acquired using pen tablet or scanner. The Krawtchouk moments are used as feature extraction technique which also invariant to shift, rotation, and scale provided that the signature pattern is in the proper format for classification. This moment method reduces the dimensionality of the input signature pattern by eliminating features containing low information or high redundancy. These features are used to recognize the signature using Fuzzy Min-Max Neural Network with Compensatory Neuron (FMCNs) which is designed by Nandelkar and Biswas.