Contactless fever measurement based on thermal imagery analysis

摘要

Fever is a very common symptom mostly in infections. As it is a reaction of the immune system, fever is supporting the healing process. However, if fever is heavily rising, it can cause serious damages to the human body. By considering the aging society user group, it is sometimes difficult to detect high fever, either due to false fever measurements, or due to the absence of a second person, in cases of single person households. Therefore, in this paper, the authors propose the design and implementation of a contactless fever measurement system, based on thermal imagery analysis. A FLIR thermal camera was used for the thermal image acquisition, interfaced with a BeagleBone Black (BBB) board, running the proposed algorithm. A platform independent solution is proposed, based on open source libraries (OpenCV, Tesseract OCR), to allow robustness in contrast to standardized branded SDKs. The proposed approach enables a compact, lightweight, hygienic and real-time fever measurement solution tool for ambient assisted living environments.