Python simulation of an IR thermometer demonstrating real-world sensor behavior including noise, drift, outliers, and signal processing techniques (moving average filter & outlier rejection).
A Python-based simulation of a Digital Infrared (IR) Thermometer for body temperature measurement. This project demonstrates how real-world environmental factors affect temperature readings over a 60-second period, including:
- Random measurement noise (Gaussian distribution)
- Temperature drift (-0.1°C/min)
- Emissivity mismatch bias
- Occasional outliers and spikes
The simulation implements signal processing techniques to improve measurement accuracy:
- Moving average filter (window size: 5)
- Statistical outlier rejection (3σ threshold)
Based on the Stefan-Boltzmann Law for infrared radiation, this project visualizes the difference between raw sensor data and processed measurements, demonstrating the importance of signal processing in real-world sensor applications.
Developed as part of the "Artificial Intelligence for Industrial Applications" program at OTH Amberg-Weiden.