학술논문

Machine Learning Techniques for Contactless Fast Body Temperature Imaging Portable Device
Document Type
Conference
Source
2024 IEEE Applied Sensing Conference (APSCON) Applied Sensing Conference (APSCON), 2024 IEEE. :1-4 Jan, 2024
Subject
Engineering Profession
General Topics for Engineers
Temperature measurement
Wireless communication
COVID-19
Wireless sensor networks
Three-dimensional displays
Imaging
Streaming media
Velocity measurement
Monitoring
Diseases
3D printing
Android application
Edge detection
Face detection
Haar Cascade
IR
Thermal Imaging
Language
Abstract
After the COVID-19 outbreak, it has been absolutely important at mass gatherings and events to check the body temperature of all the attendees before giving them entry to avoid the massive spread of diseases like COVID-19. In traditional temperature measurement methods, the time required is high, and some of them are non-hygienic due to the involvement of contact between the temperature-measuring device and the patient. So, in this paper, we have proposed a contactless, fast, portable, and safe IR-based temperature measurement system (called ThermoMudra) for simultaneously measuring the temperature of multiple people. In the proposed approach, the Haar Cascade algorithm is used for face detection in a video stream. The device is based on the thermal imaging of people, which can detect elevated temperatures to prevent contact with others. The device can be connected via hotspot to any device, and an Android app is used to monitor the temperature of people in the frame. The whole system is enclosed in a 3D printed case designed keeping in mind the self-heating of devices. It is wireless, making it a feasible portable solution for mass public gatherings. The results have been validated against a standard temperature measuring device, i.e., a digital thermometer, to understand the device’s accuracy.