Parikh Dhruv, Karthikeyan Avaneesh, Ravi V, Shibu Merin, Singh Riya, Sofana Reka S
School of Electronics Engineering, Vellore Institute of Technology, Chennai, India.
Centre for Neuroinformatics, School of Electronics Engineering, Vellore Institute of Technology, Chennai, India.
Front Public Health. 2025 May 14;13:1552515. doi: 10.3389/fpubh.2025.1552515. eCollection 2025.
COVID-19 has not only changed the way people live but has also altered the way all organizations operate. The most effective precautionary measure against the spread of the virus that caused the COVID-19 pandemic SARS-CoV-2, is to use face coverings in public settings. In this study, we present a potential application of the Internet of Things (IoT) and machine learning to prevent the spread of COVID-19. The proposed smart gateway entrance system consists of various subsystems: face mask recognition, face shield detection, face mask detection with face shields, sanitization systems, temperature monitoring systems, and vaccine verification. These systems help us to efficiently monitor, authenticate, track health parameters, and process data in real-time. The face mask and face shield detection subsystems leverage a hybrid model that combines the capabilities of MobileNetV2 and VGG19, enabling more robust and accurate detection by leveraging MobileNetV2's efficiency and VGG19's depth in feature extraction, which has an overall accuracy of 97% and notably the face shield detection component obtains an efficiency of 99%. Proposed framework includes QR code-based vaccination certificate authentication using a secure real-time database model, inspired by health platforms such as CoWIN, to ensure reliable and timely verification at points of entry and the real-time database management system developed using Haar Cascade trainer GUI helps to integrate all the data in real-time and provides access to the entry. The IoT model sanitizes individuals and tracks health parameters using an MLX90614 infrared sensor with an accuracy of ±0.5°C. As the system updates the real-time database, it helps maintain a record of the employee's health conditions and checks whether the employee follows all safety screening protocols every day. Therefore, the proposed system has immense potential to contribute to community healthcare and fight against COVID-19.
新冠疫情不仅改变了人们的生活方式，也改变了所有组织的运营方式。预防导致新冠疫情的病毒SARS-CoV-2传播的最有效预防措施，是在公共场所佩戴面部遮盖物。在本研究中，我们展示了物联网（IoT）和机器学习在预防新冠疫情传播方面的潜在应用。所提出的智能网关入口系统由多个子系统组成：口罩识别、面罩检测、佩戴面罩时的口罩检测、消毒系统、体温监测系统和疫苗验证。这些系统帮助我们高效地监测、认证、跟踪健康参数并实时处理数据。口罩和面罩检测子系统利用了一种混合模型，该模型结合了MobileNetV2和VGG19的功能，通过利用MobileNetV2的效率和VGG19在特征提取方面的深度，实现了更强大、更准确的检测，总体准确率为97%，特别是面罩检测组件的效率达到了99%。所提出的框架包括使用基于二维码的疫苗接种证书认证，采用安全的实时数据库模型，灵感来源于CoWIN等健康平台，以确保在入口处进行可靠、及时的验证，并且使用Haar Cascade训练器GUI开发的实时数据库管理系统有助于实时整合所有数据并提供入口访问权限。物联网模型使用MLX90614红外传感器对人员进行消毒并跟踪健康参数，精度为±0.5°C。随着系统更新实时数据库，它有助于记录员工的健康状况，并检查员工每天是否遵循所有安全筛查协议。因此，所提出的系统在促进社区医疗保健和抗击新冠疫情方面具有巨大潜力。
