A configurable simulator for low-resolution infrared thermal sensors: accuracy assessment and practical applications in indoor environments

Thermal sensors are increasingly used in various applications such as environmental monitoring, smart homes, and surveillance. These sensors detect infrared radiation to monitor human presence and movement, enabling sophisticated sensing capabilities. However, determining the number of sensors and optimizing their placement through real-world experimentation is often impractical due to cost and logistical constraints. This paper introduces ThermalSim, a highly configurable simulator for low-resolution thermal sensors, designed to address these issues and enable pre-testing of sensor setups in virtual environments. The simulator allows for detailed customization of environmental parameters, object properties, and sensor characteristics and supports modelling a dynamic agent with configurable trajectories and speeds. Experimental validation against real-world data demonstrates the simulator’s high accuracy in replicating static and dynamic scenarios. Metrics such as correlation, entropy and mutual information, and similarity of temperature images have been used to evaluate the simulator's output. Some case studies show the tool's flexibility, showcasing its practical applications across various scenarios.