DIGITIZED SIGNATURE FORGERY DETECTION USING HOG & SVM

Read full text Summarise with VIRAT
Authors: S Hemant Kumar, R SAI REVANTH, Y GREESHMA, U DURGA GANESH, B NAVYA
Cited by: 0
References: 11

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Handwritten signature verification is still an important part of biometric authentication systems used in banking, law, and government.This study introduces an automated offline signature verification method that use Histogram of Oriented Gradients (HOG) features in conjunction with Support Vector Machine (SVM) classification.The suggested approach solves the problems with existing manual verification methods by using strong feature extraction and machine learning techniques.HOG features record the unique edge orientations and stroke patterns of each signature.On the other hand, SVM can distinguish between actual and fraudulent samples.With a web app built on Flask, it's easy to publish and manage files safely, analyze data in real time, and have backup systems to make sure everything works.Testing shows that the classification accuracy is more than 92% on benchmark signature datasets, and the average processing time is less than one second per image.For practical use in contexts where a lot of documents need to be processed, the system architecture focuses on modularity, security, and flexibility.

Fields of Study & Subjects

Handwritten Text Recognition Techniques User Authentication and Security Systems Biometric Identification and Security

Publication details

Journal / source
International journal of engineering science and advanced technology.
Volume
26
Issue
3
Pages
518–524
Publication date
1 March 2026
Language
EN