Method of Combining the Degrees of Similarity in Handwritten Signature Authentication Using Neural Networks [chapter]

2009 Research and Development in Intelligent Systems XXVI  
This paper introduces a new method of combining the degrees of similarity for a hand-written signature (also known as holographic signature or biometric signature) using neural networks. This method is used for a biometric authentication system after the degrees of similarity between a signature and it's reference template are computed. The degrees of similarity are defined using Levenstein distance of the handwritten signature's features. Using this method we achieved the following biometric performance metrics: FRR: 8.45% and FAR: 0.9%.
doi:10.1007/978-1-84882-983-1_37 dblp:conf/sgai/Marcu09 fatcat:xg46cj6warathns7ry2w3ismwi