Human Identification Based on SIFT Features of Hand Image

Abstract

The use of hand features for human identification is a reliable and convenient method that has a greater influence than other biometric techniques because of its simplicity and dependability. In this study, we suggested a technique for identifying individuals using all the features of a hand image which differs from the previous methods of identifying individuals depending on either the features of the palmprint only or the features of the fingerprint only. Therefore, the paper aims to build an efficient system to identify individuals depending on the features of the hand image using SIFT features. Our proposed system comprises three main stages. The first stage includes removing the background and unimportant objects and extracting the hand area from the captured image. The SIFT feature is then applied to extract the robust features from the hand area. Finally, the matching stage depends on the maximum matching points of the SIFT points between the test and training image. The outcomes demonstrate the correct recognition rate (CRR), accuracy, and recall are 99.65%, 99.32%, and 98.62%, respectively. Thus, the effectiveness and efficiency of the suggested approach have been proven.