Validation and Error Characterization in UAV-Based Thermal Mapping

Thermographic measurements are becoming more and more employed for diverse applications, including agriculture and environmental monitoring. In the last decades, the effects of climate change are significantly impacting on grape phenology, yield, and wine quality, thus necessitating adaptation and conservation strategies. Unmanned Aerial Vehicle (UAV)-based thermal imaging has emerged as a promising technology in this context, enabling detailed, real-time monitoring of vineyard conditions. This study focuses on designing a methodology to validate temperature measurements from a UAV-mounted thermal sensor (DJI Zenmuse H20T) on a DJI Matrice 300 RTK platform. Data acquisition was performed in two scenarios: a controlled environment and a real-case vineyard in Montemurlo, Italy. The controlled setup assessed the sensitivity of temperature measurements to factors such as the camera-object distance, while the vineyard scenario aimed at verifying the reliability of thermal maps, checking the precision and accuracy of the estimated temperatures. This research aims to enhance the use of thermal imaging in precision agriculture, improving the accuracy of monitoring and management practices in viticulture and other related applications.