Signature Verification and Forgery Recognition System Using KNN, Backpropagation and CNN

Abstract

Signature verification is behavioral biometric and is most widely used among other types of biometric recognition system like fingerprint, voice, iris etc. There are many biometric verification systems like finger print scanning, face recognition, iris scanning, and voice recognition etc. But cost of deploying for some systems is extremely high and some of the systems are not portable. In-order to find a middle ground between efficiency, effectiveness in mass quantity and portability, we take a deeper look into signature verification system which can be deployed to produce fruitful results. In this project there is development of various algorithms to verify the signature and to analyze if it is genuine or not. It involves implementation of KNN, CNN and Backpropagation.