Application Of Neural Network In Biometric Authentication System Handwritten Signature Authentication System
Hui Yeng Yep · 2004
Handwriting signature authentication is a subset of biometric authentication. It is crucial to detect a genuine signature not only manually, but electronically as well since a lot of business transactions have been switch to electronic process lately. Traditional authentication methods face bottlenecks in guiding modern applications. Existences of biometric authentication technologies offer good alternatives to public. Over years, researchers have been presenting ideas of teaching the machines to act like human theoretically. This project seeks possibilities of applying neural network algorithm in handwriting signature authentication like bow a human does practically by creating an application to recognize human signatures. This project conducts research and attempt to build a computer program that apply neural network algorithm to recognize genuine signatures. The computer program has two interfaces that will accept input and stored data into database for future comparison. Comparison between commercial products is beyond project scope since no manufacturers have disclosed details of products' performance. Handwriting signature authentication is a good solution because this method offers moderate security level with inexpensive development and installation costs compared to other biometric technologies such as iris scans or hand geometry scans. It is a field worth attentions from academic researchers and commercial companies.