| Summary: | Solar photovoltaic (PV) panels are pivotal in renewable energy generation, yet their efficacy can be severely hampered by hotspots induced by various factors. This study introduces a pioneering approach for hotspot recognition in solar PV panels, harnessing the capabilities of the You Only Look Once (YOLO), specifically the YOLOv9 [1] model, and integrating cutting-edge image processing techniques. The aim is precise hotspot identification and localization within PV panels, facilitating targeted maintenance and optimization strategies. By amalgamating the efficiency of the YOLOv9 architecture with sophisticated image processing algorithms, the method enhances hotspot recognition performance. Extensive experimentation and validation with real-world thermal imagery datasets demonstrate the effectiveness of the approach. Results showcase substantial enhancements in hotspot detection accuracy and localization precision compared to existing methods. Moreover, the incorporation of image processing techniques streamlines targeted region identification, ensuring precise hotspot localization within PV panels. Experimental findings underscore the method's efficacy in hotspot detection across diverse environmental conditions, boosting system reliability. In essence, this research enhances solar PV panels monitoring with efficient hotspot recognition and targeted region identification, aiding proactive maintenance for plant operators. © 2024 IEEE.
|
|---|