Combined Use of Infrared Imaging and Deep-Learning Techniques for Real-Time Temperature Measurement of Train Braking Components

This is the authors' abstract. We don't add key points for this paper.

This article aims to discuss how thermal imaging could benefit from deep-learning (DL) approaches for increasing temperature measurement accuracy in those complex applications where thermal images are subject to severe temperature variations and the target component position varies continuously. Real-time temperature monitoring of braking components during train braking operations indeed represents a well-matched example, given the challenges embedded in this application case: relative camera-to-target movements, variable contrast levels in the thermal image, and so on, just to cite some. The advantage of using a specifically trained DL-based semantic segmentation model to accurately locate and track target components is demonstrated by comparing the thermal evolution of a specific component with respect to those obtained by exploiting conventional machine vision tracking methods. Experimental validation in a high-speed railway environment shows the proposed method reduces temperature bias to less than 0.5 °C when compared to thermocouple data. This accuracy is crucial for ensuring the safety and reliability of railway braking systems. While the widespread use of thermal cameras on trains may be impractical, the approach demonstrates that it is possible to effectively exploit semantic segmentation models to enhance thermal imaging and retrieve more accurate temperature data in complex real-time applications.