Forgery Detection In Handwritten Documents
S Anandhamurugan, D Prasanth, R. Angel Sujitha, N. Mukhilan · 2024
This article introduces a novel method for handwritten document forgery detection based on Siamese neural networks. By training on pairings of authentic and forged document samples, the network learns to extract discriminative properties for distinguishing real from counterfeit instances. Our approach achieves higher accuracy and processing efficiency in experimental evaluations conducted on a benchmark dataset compared to both state-of-the-art and conventional approaches. The proposed Siamese neural network shows promise in automatically identifying forgeries, offering a dependable solution applicable to a wide range of fields requiring document verification and authentication. This research advances the field of document forensics and provides a pathway to enhanced security and reliability by leveraging deep learning techniques to solve the challenges in handwriting forgery detection.