A Comprehensive Review of Thermal-Aware Face Recognition Systems: Progress and Challenges

Abstract

Modern biometric systems have relied on face recognition as it is precise, convenient and unobtrusive. However, standard visible-spectrum face recognition systems are very much sensitive to changes in lighting, facial expressions, and surroundings and at the same time have no ability to conduct physiological measurements in tandem. Within the framework of access control and overall safety needs in the post-pandemic context, these restrictions have indicated a gap in research that is urgent: the lack of integrated frameworks that carry out the functions of identity verification and health-related screening in tandem with each other. This article is a critical survey of the state-of-the-art in the field of thermal-sensitive face recognition, which integrates the visible RGB imaging with biometric recognition and long-wave infrared (LWIR) thermal sensing with the purpose of estimating body temperature. We syntactically examine the available architectures, sensing configurations, fusion strategies, datasets, and evaluation protocols found in the literature. The review emphasizes the fact that dual-modal systems have the potential to allow real-time and contactless identity verification and support a large-scale approach in high-traffic settings. Moreover, this paper addresses the main technical and practical issues that are limiting the large-scale implementation, such as sensor calibration, cross-modular data alignment, environmental bias, data privacy, as well as ethical factors. Last but not the least, we describe the new research directions and future outlooks of the unified biometric and health-conscious access control systems through a critical lens that analyzes performance trade-offs, computational costs, and ethical implications, with the aim of informing the researchers and practitioners of the creation of effective, scalable, and privacy-aware solutions.