International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 05 Issue: 05 | May-2018 www.irjet.net p-ISSN: 2395-0072 © 2018, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 4206 Offline Signature Verification Using Neural Network Dr. Kiran Y.C 1 , Ms. Nirmita Nagaraj 2 1 Professor, Dept. of Computer Science and engineering, B.N.M Institute of Technology, Karnataka, India 2 Student, Dept. of Computer Science and engineering, B.N.M Institute of Technology, Karnataka, India ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - — Signature is a distinct or a characteristic of a person’s name as a source of identification. Signature is required to approve, accept or to oblige anything by the individual on the document. Hence it considered as one of the Biometric. These days, signatures are pruned to forgeries hence, an effective offline signature verification using neural network as a classifier is proposed in this paper. The test signature is identified first to see who the signer is and then the test signature undergoes the verification process to check if it is a forged or a genuine signature. Once the image is uploaded into the system, it undergoes pre-processing and feature extraction. The trained result is sent to the neural classifier along with the extracted features and the verification process is done to obtain the final result. The output is shown in the form of matched percentage. The performance of the model shows that it works efficiently with fewer number of images. Key Words: Biometric, offline signature verification, neural network, test signature, trained result. 1. INTRODUCTION One of the distinguishing feature for a person’s identification through many ages has been signature. Signature is one of the Biometric for person’s identity. A person’s name or a simple short form of the name can be a handwritten depiction that acts as a proof of identification. One of the accepted form of conformation, be it in the form of transactions, document verification, contracts in civil law, voting or in authenticating one’s identity all of them are authorized via signature. If authenticity or verification has to be done on a regular basis, then it has to happen automatically. Hence automatic signature verification comes into picture. This fastens the speed of verification automatically. One of the best application of signature is financial application and border control. Handwritten signatures are considered as a proof of identity for legal documents such as bank cheques, legal property documents, etc. Signature is considered as one of the behavioral form of human characteristic. Signature identification and verification are well established and is efficient. Biometrics are widely used these days as it is considered as one of the safest and most unique form of recognizing a person’s identity. Machine Learning is one of the paradigm that is used in Biometrics. Biometrics is considered as one of the traditional form of an individual’s authentication. Traditional methods are becoming more obsolete these days and are no longer used. ID cards and passwords are gradually taken away by Biometrics in this generation. More the metrics given for identification greater is the accuracy. In this case, if the characteristic’s given for verification is multimodal, then it is almost nearly impossible to forge. It provides an identification that is unique in nature. This way of identifying humans by their traits is referred to as Biometric Authentication. In general, Biometrics are automated methods of recognizing a person based on physiological or behavioral characteristics. Biometric management systems consist of hardwares and softwares which is not complicated to analyze and install. The process of installation is not time consuming rather it can be easily installed. For years, handwritten signatures have been a potential way of accessing one’s identity. The identification is made at ease to any individual working in this approach. Handwritten signatures are proved to be more evident than any other form of Biometric. The person signing the document is called a Signer or a Signatory. Biometrics is widely used in security applications. Since technology is increasing these days, so is the threat in accessing an individual’s identity is increasing. Hence the security measure in identification and verification is adopted. Ever since decades, this has been an open research problem. Various techniques and methods have been adopted to resolve this problem. Signatures are in the form of two types:  Physiological  Behavioral Physiological signatures are in the form of iris, cortex, thumb impression, face recognition. The physical feature is directly captured in this form. Behavioral signatures are Biometrics in the form of handwritten signatures, voice recognition. Here two