Digital signature of document using human palm

Abstract

A digital signature is a checksum which depends on the time period during which it was produced. Human palm biometric is one of the fastest, accurate, reliable and secure biometric techniques for identification and verification because it provides automatic authentication of an individual based on unique features in palm structure. In this paper, an efficient digital signature model for the document is proposed by using human palm. Human palm can give unique features which can be used in generating a secure digital signature. Therefore, this model consists of two sides: the embedding side and extracting side. The embedding side includes (1) image preprocessing stage:(color to grayscale and histogram equalization). (2) feature extraction stage:(GLCM (Haralick) algorithm). (3) Generating digital signature stage:(Elliptic Curve and Cubic Spline function with MD5 algorithm). While the extracting side contains extracting signature stage and matching stage. The accuracy of the generated digital signature by the proposed model is 100%, however false acceptance rate (FAR) is 0%, false reject rate (FRR) is 0%, and equal error rate (ERR) is 0%.