“FINGERPRINT MATCHING USING NEURAL NETWORK TRAINING”

Abstract

Fingerprint is widely accepted for personnel identification. In this paper Levenberg-Marquardt back propagation (LMBP) algorithm is used for training purpose because LMBP is fastest technique for complex data sets and gives better performance in such situation. Input image is trained by trainlm function for producing different result sets like performance plot, regression values, simulation network of input image , histogram graph etc