Evaluation of Faults in a Photovoltaic Power Plant using Orthomosaics based on Aerial Infrared Thermography

This paper aims to analyze different methods and software for the orthomosaic reconstruction of aerial infrared thermography (aIRT) images of a 3 MW PV plant in Brazil. Results showed that the method could improve the process of inspections of PV plants through aerial imagery, facilitating the localization of faults. Some challenges observed include the long duration of the flights caused by the high overlapping of images required, and the distortions presented in the resulting orthomosaics. The combination of the method with artificial intelligence algorithms that detect faults in aIRT images is a fast and effective tool to inspect PV power plants.