期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2016
卷号：4
期号：4
页码：6544
DOI：10.15680/IJIRCCE.2016.0404037
出版社：S&S Publications
摘要：The signature of a person is a biometric trait of a human being which is used to confirm human identity. A number of methods have been recommended for the identification of signatures. There are two kinds of authentication systems, o ne is the vision based and the other one is the non - vision based. The vision based includes face recognition, fingerprint recognition, iris scanning and retina scanning and the non - vision based includes voice recognition and signature verification. M ajority of the organizations primarily focus on the visual appearance of the signature for verification purposes. Many official papers, such as bonds, cheques, and card transactions require the signing of a signature. We need an automatic verification syst em because of the adverse side - effect of being easily abused by those who would fake the identification or objective of an individual. An offline signature verification system generally consists of four components: data attainment, processing, extraction of features, recognition and verification. In this system t he signature is captured and presented to the user in the form of an image. The Signatures are then verified based on parameters extracted from the signature using different image processing techniqu es. The system was tested with signatures including both genuine and forged signatures and then letting the system decide whether the signature is genuine or forged. In this paper we verify whether the signature using the neural network architecture
关键词：Neural Network; Signature Verification and Back Propagation Algorithm
