Fingerprint Authentication using Shark Smell Optimization Algorithm

• Main Authors: Bakhan Tofiq Ahmed, Omar Younis Abdulhameed
• Format: Article
• Language: English
• Published: University of Human Development 2020-07-01
• Series: UHD Journal of Science and Technology
• Subjects: fingerprint authentication; feature extraction; swarm intelligent; shark smell optimization; zkt eco device
• Online Access: http://journals.uhd.edu.iq/index.php/uhdjst/article/view/747/549

Recognition of people relying on biometric characteristics is a common phenomenon in our society. It has increased in recent years in most areas of life such as government, department, companies, and banks. Fingerprint identification is one of the most common and credible personal biometric identification methods. Extracting features are one of the most important steps in the fingerprint identification; the strength of any system depends mainly on this step, where whenever the features obtained are accurate whenever the identification process is more accurate. Therefore, an effective and efficient method must be used to extract the features. This paper solved two main problems that were (1) improving security by designing and implementing an accurate, efficient, and fast authentication system for the identification and verification process using an intelligent algorithm to extract the best features from the fingerprint image and (2) evaluating the strength of the Shark Smell Optimization (SSO) in the search space with a chosen set of metrics. This paper aims to extract the best features of the fingerprint image using an algorithm that depends on nature for its movement and work; therefore, the SSO was used. In this paper, the SSO algorithm is used to extract the features. SSO is a new meta-heuristic algorithm that uses smart methods and random movements to get its prey. These methods and movements were used to extract features from the fingerprint image which will be used later for identification and verification process. The proposed method is implemented through four phases, namely, create a database to store and organize data, image pre-processing using median filter, feature extraction using SSO algorithm, and matching process using euclidean distance. The results demonstrated the strength, accurate, credible, and effectiveness of the algorithm used by applying it on (150) real fingerprint samples taken from university students, where the results of false acceptation rate, false rejection rate, and correct verification rate were 0.00, 0.00666, and 99.334%, respectively.