Automated Signature Verification based on Hybrid Features and Proposed Deep Model

Signatures are behavioral biometric traits of a person, used to authenticate a person. In all the legal transactions and legal documents signature is required to authenticate its legality. In such cases there are chances to forge the signature by other person to get the benefits. Therefore in order to check the genuineness of the signature, signature verification system is needed. In the state-of-the art literature there are several algorithms are proposed by different authors but still few challenges are remained to address such as the detection of skilled forgery and detection of intra class variations. There are two types of signature verification system namely offline and online. For such authentication of signature this project presents an application software which facilitates the feature of offline signature verification using the convolution neural network approach. This software is able to train the network with new dataset of signature and validate the authenticity of new signature of trained class. User can also perform experiment and analyze the training and verification of model with features result analysis feature. This software consists very efficient user interface so that any non- technical person can use the software without any difficulty. The aim of signature verification is to discriminate genuine signature from forge signature. Forgeries are of three types first one is simple forgery, second one is random forgery and third one is skilled forgery. This software consists very efficient user interface so that any non- technical person can use the software without any difficulty. For measuring the performance of system, the neural network was trained with dataset of 27 person with each having at least 100 signature and training result was 97 % accuracy.