Book Section #16957

Abstract

Biometrics are a branch of science that is used to identify as well as authenticate. Biometrics are basically of two types: behavioral biometrics and physiological biometrics. Characteristics and biometrics are fundamentally fixed and unique, allowing individuals to distinguish one from another. Biometric authentication systems have received more attention in recent years than other traditional authentication methods such as passwords or signatures. All human biological traits are unique as biometrics such as fingerprints, palms, irises, palm blood vessels and fingerprint blood vessels, and other biometrics. Biometric identification systems basically have a complex structure that consists of different parts. Biometric-based authentication systems and authentication methods, along with other authentication systems, can improve the security aspects of authentication systems. Identification methods and tools are used in many important and essential applications such as surveillance processes, security investigations, fraud detection technologies, and access controls. Biometric-based identification methods in machine learning consist mainly of preprocessing, feature extraction, feature selection, classification, and finally evaluation. These systems can also be based on one biometric or based on several biometrics together. In this chapter, we examine the methods of identifying identity information with the help of various biometrics, highlight the challenges in each biometrics, and introduce the solutions that have been proposed to overcome this challenge. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.