Scholarlink

Nowadays, the use of unmanned aerial vehicles (UAVs) for aerial inspections of photovoltaic (PV) plants has emerged as a crucial alternative solution. This study introduces a method for detecting PV modules and arrays using RGB images processed with YOLOv8. A dataset of RGB images was collected using a drone (DJI mini-3) at the Unit for Solar Equipment Development (UDES), located in northern Algeria, and employed to develop this method. Beyond the detection of PV modules and PV panels, this approach also helps users to detect, classify and locate certain types of PV systems faults and thus, determine the need to launch a maintenance procedure based on visual inspections. Detailed RGB images collected. The results are promising, with a detection accuracy ranging between 70% and 95%. Based on the experience gained, the potential and limitations of the developed method for application to larger-scale PV plants are also discussed.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

YOLOv8-Based Photovoltaic Module Detection Using Aerial Imagery

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • N. Kellil,
  • C. Moussaoui,
  • A. Mellit,
  • A. Boudiaf

摘要

Nowadays, the use of unmanned aerial vehicles (UAVs) for aerial inspections of photovoltaic (PV) plants has emerged as a crucial alternative solution. This study introduces a method for detecting PV modules and arrays using RGB images processed with YOLOv8. A dataset of RGB images was collected using a drone (DJI mini-3) at the Unit for Solar Equipment Development (UDES), located in northern Algeria, and employed to develop this method. Beyond the detection of PV modules and PV panels, this approach also helps users to detect, classify and locate certain types of PV systems faults and thus, determine the need to launch a maintenance procedure based on visual inspections. Detailed RGB images collected. The results are promising, with a detection accuracy ranging between 70% and 95%. Based on the experience gained, the potential and limitations of the developed method for application to larger-scale PV plants are also discussed.