DRONE-BASED SOLAR PANEL FAULT CLASSIFICATION USING INFRARED IMAGES AND DEEP LEARNING: A COMPARATIVE STUDY OF A LIGHTWEIGHT CNN AND RESNET18
##plugins.themes.academic_pro.article.main##
Abstract
Efficient monitoring of photovoltaic (PV) modules is essential for sustaining
energy yield and reducing manual inspection effort in solar power plants. Dronebased
infrared (IR) imaging is a practical inspection modality because thermal
anomalies often indicate electrical or physical faults