The accumulation of dust, dirt, bird droppings, pollen, and other environmental contaminants on photovoltaic (PV) panels significantly reduces their power generation efficiency and increases operational and maintenance costs. Therefore, timely and accurate detection of the cleanliness condition of solar panels is essential for maintaining optimal energy production, improving system reliability, and extending the operational lifespan of PV installations. In recent years, considerable research interest has been directed toward automated soiling detection techniques, particularly those based on image processing, computer vision, and artificial intelligence. This paper presents a comprehensive survey of existing methods for detecting and classifying clean and contaminated PV panels, with a primary focus on vision- based deep learning approaches. Initially, conventional soiling detection methods, including sensor-based and power-output-based techniques, are reviewed along with their limitations, such as high implementation cost, limited scalability, and sensitivity to environmental variations. The survey then discusses recent advancements in image processing and deep learning, including feature extraction techniques, transfer learning strategies, and efficient neural network architectures such as YOLO, EfficientNet, and MobileNet. YOLO-based models are particularly emphasized because of their high detection accuracy, fast inference speed, and suitability for real-time and edge-based applications. Furthermore, this review examines commonly used datasets, image pre-processing techniques, and evaluation metrics, including accuracy, precision, recall, F1-score, confusion matrix, and ROC–AUC. A comparative analysis of published studies indicates that deep learning-based methods consistently outperform conventional approaches by providing higher accuracy, greater robustness, and improved adaptability under diverse environmental conditions. Finally, the paper discusses key challenges, including dataset imbalance, varying soiling severity, illumination changes, and geographical diversity, while highlighting future research directions for developing reliable, efficient, and intelligent PV panel soiling detection systems.

Keywords - Solar panel soiling, Photovoltaic monitoring, Clean/dirty classification, Deep learning, YOLO, Image-based detection, Computer vision, ROC–AUC, Smart maintenance.

🔒 This is a subscription article

Full text is available to subscribers and institutional members. Please choose an option below to access it.

1. Kazem AA, Chaichan MT, Kazem HA. Dust effect on photovoltaic utilization in Iraq. Renewable and Sustainable energy reviews. 2014 Sep 1;37:734-49.
2. Mani M, Pillai R. Impact of dust on solar photovoltaic (PV) performance: Research status, challenges and recommendations. Renewable and sustainable energy reviews. 2010 Dec 1;14(9):3124-31.
3. Ilse K, Micheli L, Figgis BW, Lange K, Daßler D, Hanifi H, Wolfertstetter F, Naumann V, Hagendorf C, Gottschalg R, Bagdahn J. Techno-economic assessment of soiling losses and mitigation strategies for solar power generation. Joule. 2019 Oct 16;3(10):2303-21.
4. Alkharusi T, Alzahrani MM, Pandey C, Yildizhan H, Markides CN. Experimental investigation of nonuniform PV soiling. Solar Energy. 2024 Apr 1;272:112493.
5. Mejia FA, Kleissl J. Soiling losses for solar photovoltaic systems in California. Solar energy. 2013 Sep 1;95:357-63.
6. Abdelsattar M, Abdelmoety A, Ismeil MA, Emad-Eldeen A. Automated defect detection in solar cell images using deep learning algorithms. Ieee access. 2025 Jan 3;13:4136-57.
7. Deitsch S, Christlein V, Berger S, Buerhop-Lutz C, Maier A, Gallwitz F, Riess C. Automatic classification of defective photovoltaic module cells in electroluminescence images. Solar Energy. 2019 Jun 1;185:455-68..
8. Tahir MF, Lamichhane S, Tzes A, Fang Y, El-Fouly TH, Umar S. Optimizing dust accumulation quantification on photovoltaic panels using deep learning visual models with hyperparameter optimization. Renewable Energy. 2025 Oct 1;251:123440.
9. Matusz-Kalász D, Bodnár I, Jobbágy M. An overview of CNN-based image analysis in solar cells, photovoltaic modules, and power plants. Applied Sciences. 2025 May 14;15(10):5511.
10. Redmon J, Farhadi A. Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767. 2018 Apr 8.
11. Bochkovskiy A, Wang CY, Liao HY. Yolov4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934. 2020 Apr 23.
12. Yaseen M. What is YOLOv9: An in-depth exploration of the internal features of the next-generation object detector. arXiv preprint arXiv:2409.07813. 2024 Sep 12.
13. Tan M, Le Q. Efficientnet: Rethinking model scaling for convolutional neural networks. InInternational conference on machine learning 2019 May 24 (pp. 6105-6114). PMLR.
14. Fawcett T. An introduction to ROC analysis. Pattern recognition letters. 2006 Jun 1;27(8):861-74.
15. Sokolova M, Lapalme G. A systematic analysis of performance measures for classification tasks. Information processing & management. 2009 Jul 1;45(4):427-37.
16. Mellit A, Kalogirou SA, Hontoria L, Shaari S. Artificial intelligence techniques for sizing photovoltaic systems: A review. Renewable and Sustainable Energy Reviews. 2009 Feb 1;13(2):406-19.
17. Sepúlveda-Oviedo EH, Estibals B. Framework for effective PV system instrumentation focused on fault diagnosis. Solar Energy Advances. 2025 Jul 14;5:100112.
18. Kimber A, Mitchell L, Nogradi S, Wenger H. The effect of soiling on large grid- connected photovoltaic systems in California and the southwest region of the United States. In2006 IEEE 4th World Conference on Photovoltaic Energy Conference 2006 May 7 (Vol. 2, pp. 2391-2395). IEEE.
19. Jiang H, Lu L, Sun K. Experimental investigation of the impact of airborne dust deposition on the performance of solar photovoltaic (PV) modules. Atmospheric environment. 2011 Aug 1;45(25):4299-304.
20. Sayyah A, Horenstein MN, Mazumder MK. Energy yield loss caused by dust deposition on photovoltaic panels. Solar Energy. 2014 Sep 1;107:576-604.
21. Buerhop C, Weißmann R, Scheuerpflug H, Auer R, Brabec CJ. Quality control of PV- modules in the field using a remote-controlled drone with an infrared camera. InProceedings of the 27th European Photovoltaic Solar Energy Conference and Exhibition (EU PVSEC 2012), Frankfurt, Germany 2012 Sep 24 (Vol. 1, pp. 3370-3373).
22. He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. InProceedings of the IEEE conference on computer vision and pattern recognition 2016 (pp. 770-778).
23. Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks. Communications of the ACM. 2017 May 24;60(6):84-90.
24. Lin TY, Goyal P, Girshick R, He K, Dollár P. Focal loss for dense object detection. InProceedings of the IEEE international conference on computer vision 2017 (pp. 2980- 2988).
25. Chen T, Guestrin C. Xgboost: A scalable tree boosting system. InProceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining 2016 Aug 13 (pp. 785-794).
26. Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X, Unterthiner T, Dehghani M, Minderer M, Heigold G, Gelly S, Uszkoreit J. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929. 2020 Oct 22.
27. Woo S, Park J, Lee JY, Kweon IS. Cbam: Convolutional block attention module. InProceedings of the European conference on computer vision (ECCV) 2018 (pp. 3-19).
28. Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation. InProceedings of the IEEE conference on computer vision and pattern recognition 2015 (pp. 3431-3440).
29. Han S, Mao H, Dally WJ. Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding. arXiv preprint arXiv:1510.00149. 2015 Oct 1.
30. Howard AG, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T, Andreetto M, Adam H. Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861. 2017 Apr 17.

The accumulation of dust, dirt, bird droppings, pollen, and other environmental contaminants on photovoltaic (PV) panels significantly reduces their power generation efficiency and increases operational and maintenance costs. Therefore, timely and accurate detection of the cleanliness condition of solar panels is essential for maintaining optimal energy production, improving system reliability, and extending the operational lifespan of PV installations. In recent years, considerable research interest has been directed toward automated soiling detection techniques, particularly those based on image processing, computer vision, and artificial intelligence. This paper presents a comprehensive survey of existing methods for detecting and classifying clean and contaminated PV panels, with a primary focus on vision- based deep learning approaches. Initially, conventional soiling detection methods, including sensor-based and power-output-based techniques, are reviewed along with their limitations, such as high implementation cost, limited scalability, and sensitivity to environmental variations. The survey then discusses recent advancements in image processing and deep learning, including feature extraction techniques, transfer learning strategies, and efficient neural network architectures such as YOLO, EfficientNet, and MobileNet. YOLO-based models are particularly emphasized because of their high detection accuracy, fast inference speed, and suitability for real-time and edge-based applications. Furthermore, this review examines commonly used datasets, image pre-processing techniques, and evaluation metrics, including accuracy, precision, recall, F1-score, confusion matrix, and ROC–AUC. A comparative analysis of published studies indicates that deep learning-based methods consistently outperform conventional approaches by providing higher accuracy, greater robustness, and improved adaptability under diverse environmental conditions. Finally, the paper discusses key challenges, including dataset imbalance, varying soiling severity, illumination changes, and geographical diversity, while highlighting future research directions for developing reliable, efficient, and intelligent PV panel soiling detection systems.

Keywords - Solar panel soiling, Photovoltaic monitoring, Clean/dirty classification, Deep learning, YOLO, Image-based detection, Computer vision, ROC–AUC, Smart maintenance.

🔒 This is a subscription article

Full text is available to subscribers and institutional members. Please choose an option below to access it.

1. Kazem AA, Chaichan MT, Kazem HA. Dust effect on photovoltaic utilization in Iraq. Renewable and Sustainable energy reviews. 2014 Sep 1;37:734-49.
2. Mani M, Pillai R. Impact of dust on solar photovoltaic (PV) performance: Research status, challenges and recommendations. Renewable and sustainable energy reviews. 2010 Dec 1;14(9):3124-31.
3. Ilse K, Micheli L, Figgis BW, Lange K, Daßler D, Hanifi H, Wolfertstetter F, Naumann V, Hagendorf C, Gottschalg R, Bagdahn J. Techno-economic assessment of soiling losses and mitigation strategies for solar power generation. Joule. 2019 Oct 16;3(10):2303-21.
4. Alkharusi T, Alzahrani MM, Pandey C, Yildizhan H, Markides CN. Experimental investigation of nonuniform PV soiling. Solar Energy. 2024 Apr 1;272:112493.
5. Mejia FA, Kleissl J. Soiling losses for solar photovoltaic systems in California. Solar energy. 2013 Sep 1;95:357-63.
6. Abdelsattar M, Abdelmoety A, Ismeil MA, Emad-Eldeen A. Automated defect detection in solar cell images using deep learning algorithms. Ieee access. 2025 Jan 3;13:4136-57.
7. Deitsch S, Christlein V, Berger S, Buerhop-Lutz C, Maier A, Gallwitz F, Riess C. Automatic classification of defective photovoltaic module cells in electroluminescence images. Solar Energy. 2019 Jun 1;185:455-68..
8. Tahir MF, Lamichhane S, Tzes A, Fang Y, El-Fouly TH, Umar S. Optimizing dust accumulation quantification on photovoltaic panels using deep learning visual models with hyperparameter optimization. Renewable Energy. 2025 Oct 1;251:123440.
9. Matusz-Kalász D, Bodnár I, Jobbágy M. An overview of CNN-based image analysis in solar cells, photovoltaic modules, and power plants. Applied Sciences. 2025 May 14;15(10):5511.
10. Redmon J, Farhadi A. Yolov3: An incremental improvement. arXiv preprint arXiv:1804.02767. 2018 Apr 8.
11. Bochkovskiy A, Wang CY, Liao HY. Yolov4: Optimal speed and accuracy of object detection. arXiv preprint arXiv:2004.10934. 2020 Apr 23.
12. Yaseen M. What is YOLOv9: An in-depth exploration of the internal features of the next-generation object detector. arXiv preprint arXiv:2409.07813. 2024 Sep 12.
13. Tan M, Le Q. Efficientnet: Rethinking model scaling for convolutional neural networks. InInternational conference on machine learning 2019 May 24 (pp. 6105-6114). PMLR.
14. Fawcett T. An introduction to ROC analysis. Pattern recognition letters. 2006 Jun 1;27(8):861-74.
15. Sokolova M, Lapalme G. A systematic analysis of performance measures for classification tasks. Information processing & management. 2009 Jul 1;45(4):427-37.
16. Mellit A, Kalogirou SA, Hontoria L, Shaari S. Artificial intelligence techniques for sizing photovoltaic systems: A review. Renewable and Sustainable Energy Reviews. 2009 Feb 1;13(2):406-19.
17. Sepúlveda-Oviedo EH, Estibals B. Framework for effective PV system instrumentation focused on fault diagnosis. Solar Energy Advances. 2025 Jul 14;5:100112.
18. Kimber A, Mitchell L, Nogradi S, Wenger H. The effect of soiling on large grid- connected photovoltaic systems in California and the southwest region of the United States. In2006 IEEE 4th World Conference on Photovoltaic Energy Conference 2006 May 7 (Vol. 2, pp. 2391-2395). IEEE.
19. Jiang H, Lu L, Sun K. Experimental investigation of the impact of airborne dust deposition on the performance of solar photovoltaic (PV) modules. Atmospheric environment. 2011 Aug 1;45(25):4299-304.
20. Sayyah A, Horenstein MN, Mazumder MK. Energy yield loss caused by dust deposition on photovoltaic panels. Solar Energy. 2014 Sep 1;107:576-604.
21. Buerhop C, Weißmann R, Scheuerpflug H, Auer R, Brabec CJ. Quality control of PV- modules in the field using a remote-controlled drone with an infrared camera. InProceedings of the 27th European Photovoltaic Solar Energy Conference and Exhibition (EU PVSEC 2012), Frankfurt, Germany 2012 Sep 24 (Vol. 1, pp. 3370-3373).
22. He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. InProceedings of the IEEE conference on computer vision and pattern recognition 2016 (pp. 770-778).
23. Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks. Communications of the ACM. 2017 May 24;60(6):84-90.
24. Lin TY, Goyal P, Girshick R, He K, Dollár P. Focal loss for dense object detection. InProceedings of the IEEE international conference on computer vision 2017 (pp. 2980- 2988).
25. Chen T, Guestrin C. Xgboost: A scalable tree boosting system. InProceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining 2016 Aug 13 (pp. 785-794).
26. Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X, Unterthiner T, Dehghani M, Minderer M, Heigold G, Gelly S, Uszkoreit J. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929. 2020 Oct 22.
27. Woo S, Park J, Lee JY, Kweon IS. Cbam: Convolutional block attention module. InProceedings of the European conference on computer vision (ECCV) 2018 (pp. 3-19).
28. Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation. InProceedings of the IEEE conference on computer vision and pattern recognition 2015 (pp. 3431-3440).
29. Han S, Mao H, Dally WJ. Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding. arXiv preprint arXiv:1510.00149. 2015 Oct 1.
30. Howard AG, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T, Andreetto M, Adam H. Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861. 2017 Apr 17.