Computer Vision-Based Signature Forgery Detection System Using Deep Learning: A Supervised Learning Approach
Ryan Carreon Reyes, Myriam J. Polinar, Richardson M. Dasalla, Godofredo S. Zapanta, Mark P. Melegrito, Renato Racelis Maaliw III · 2022 IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT) · 2022
Authentication is a crucial aspect of data security. It is one of the most important issues of our time. As technology advances, our interactions with machines are becoming increasingly automated. As a result, for a variety of security concerns, the demand for authentication is rapidly expanding. As a result, biometric-based authentication has become extremely popular. It has a significant edge over other approach. However, because different ways are utilized to verify people, this incidence is not a substitute for a problem. Signatures were one of the first commonly utilized biometric traits for identifying people. This paper describes a method for simplifying signature verification by preprocessing signatures. It also included a novel deep learning-based method for detecting faked signatures. With an accuracy of 85-95 %, the proposed method detects forgeries.