Monitoring temperature-related hazards using mobile devices and a thermal camera

Abstract

The paper proposes an algorithm that allows for automatic detection of thermal hazards using a smartphone-type device and a mobile thermal camera. The algorithm works in two stages. First, it analyses the environment of a disabled person for areas with a high or very high temperature. If they are detected, it signals by voice messages the appearance of such an object in the camera field of view. The second stage is the detection of a situation when a visually impaired person brings their hand closer to a hot object. In this situation, the application automatically detects the hand and the hot object and signals by voice messages a higher hazard level. In order to classify the level of thermal hazards, image thresholding methods were used, which enable to detect objects with temperatures above 40, 55, 70 $$^\circ $$C. For hand detection in a situation threatening burns, an approach based on sequential thermal image analysis and machine learning was used. Tests were carried out for various algorithm configurations. The accuracy and precision values were as follows: Acc $$=0.89$$, Prec $$=0.91$$. The proposed method allows for quick and completely automatic warning of hazards and can be used for popular Android mobile devices.