Classification of Static Signature Based on Distance Measure Using Feature Selection

Abstract

In this paper, eigenvector-based moments are proposed for offline signature validation. Here, principal component analysis (PCA) and linear discriminant analysis (LDA) techniques are used for dimension reduction and generated eigenvector which is calculated using Euclidean distance. It measured the distance between two vectors having an equal size in 2-D space. A newly suggested approach to generate the eigenvector from training and testing samples of signatures, which is calculated through Euclidean distance as a classifier. In which, it has shown high verification accuracy of 91.07% on the MCYT-75 corpus and GPDS synthetic signature database.