Article Preview
TopIntroduction
Access control system (Zhang et al. 2018) is an important part of modern enterprise management information system, which provides a basis for effectively evaluating user attendance and enterprise security. The access control system based on biometric features can effectively avoid the problems of forgery, loss, theft or forgetting in traditional attendance methods such as manual check-in, magnetic card and IC card, and ensure the authenticity and effectiveness of attendance data.
With the continuous development of society, there is an increasing demand for fast, effective and non-invasive automatic identity authentication technology in all aspects of society, especially in the field of security. Traditional identity authentication technology mainly depends on personal identity information card (such as ID card, bank card) or password account number. Such methods are not only inconvenient to carry and easy to lose, but also easy to damage, crack and steal. In order to solve the above problems, many identity authentication methods based on biometric technology have been widely proposed.
At present, the biometric recognition technologies mainly include fingerprint recognition (Cao & Jain 2018), iris recognition (Zhao & Kumar 2017), retinal recognition (Rajan 2020), face recognition (Deng et al. 2021) and so on. Compared with the traditional identification features such as ID card and password, the biological features of human individuals have stronger independence and stability, and are difficult to forge, damage and lose. Therefore, automatic identity authentication technology based on biometric recognition technology (fingerprint, iris, face, retinal etc.) has a wide application prospect and theoretical research value.
Among many biometric based authentication technologies, face recognition technology (Masi et al. 2018; Adjabi et al. 2020; Kortli et al. 2020) has become a research hotspot in the field of biometric recognition due to direct, friendly, convenient, hidden and efficient characteristics of face features. Compared with other biometric recognition technologies, such as DNA recognition (Xu et al. 2018), iris recognition (Ngyyen et al. 2017) and fingerprint recognition (Prasad et al. 2018), the feature extraction process of face recognition is more convenient. For example, iris recognition and fingerprint recognition need to accurately align specific parts of the body with the corresponding hardware for recognition, while DNA recognition needs to obtain other people's blood, hair and other items. Face recognition can break away from these constraints. In general visual situations, face images can be captured normally. At the same time, face recognition technology has low cost and requirements for hardware and surrounding environment.
At the same time, face recognition technology is simple to operate and has strong post tracking ability. The people can be verified and identified by face images. During the process of person identification and authentication, the non-contactable feature extraction method is more welcome. In the field of security, face recognition technology does not need the cooperation of users and has excellent concealment. It is very suitable for practical applications such as security monitoring and case detection. These factors determine that face recognition technology has a wide application prospect.
The main applications of face recognition include: case detection (Hartanto & Adji 2018), video surveillance (Jamil et al. 2017), identity authentication (Liu et al. 2019) etc. In the case detection, the face recognition can quickly locate or narrow the scope of criminals by comparing the intercepted face images in video surveillance with the face database inside the public security system to help detect cases quickly and effectively. In video surveillance, the face detection and face recognition for the surveillance video can give an alarm when a specific person bursts into the monitoring range. In identity authentication, the face recognition can replace the traditional access control authentication technology, such as traditional keys, authentication cards and password.