An Ensemble-Based Deep Learning Framework for Efficient Soiling Detection on Photovoltaic Panels

Year 2026, Volume: 6 Issue: 1, 1 - 12, 27.02.2026

Musa Balcı , Andaç Fındıkçı , Mustafa Yasin Erten , Hüseyin Aydilek

https://doi.org/10.5152/tepes.2025.25020

https://izlik.org/JA45GX56DS

Abstract

Solar energy plays a pivotal role in renewable energy systems; however, dust accumulation on photovoltaic panels substantially reduces energy production efficiency. Manual cleaning methods at large-scale plants are costly and impractical, highlighting the need for automated detection techniques. This study presents a novel image processing and deep learning-based approach to accurately detect dusty PV panels. Images underwent preprocessing, including Hue, Saturation, Value color space conversion, and morphological operations to precisely segment dust-affected regions. Individual performances of DenseNet169, Xception, and InceptionV3 models were evaluated, and an ensemble model—Deep Solar Ensemble—was developed via soft voting. Experimental results demonstrated that the proposed ensemble achieved a superior classification accuracy of 97.02%, a precision of 97.29%, a recall of 96.56%, and an F1 score of 96.92% on a binary classification task. To address real-world applicability and robustness, the study was extended to include comparisons with lightweight architectures and testing on a more diverse, multi-class dataset containing various fault types, where the ensemble continued to show robust performance. The proposed methodology offers significant potential for automating solar panel maintenance, thereby enhancing operational efficiency, while also considering the tradeoffs between accuracy and computational cost for practical deployment.

Keywords

DenseNet , dust classification , ensemble learning , InceptionV3 , photovoltaic panel , Xception

References

• 1. Ç. D. Gezgin, “Güneş panellerinde, güneş takip sistemlerinin ve panel kirliliğinin panel verimliliğine etkisinin incelenmesi,” M.S. thesis, Dept. of Mechatronics Eng., Inst. of Sci., Trakya Univ., Edirne, Turkey, 2023
• 2. R. Aman, M. Rizwan, and A. Kumar, “Fault classification using deep learning based model and impact of dust accumulation on solar photovoltaic modules,” Energy Sources A, vol. 45, no. 2, pp. 4633–4651, 2023.
• 3. M. Onim et al., “SolNet: A convolutional neural network for detecting dust on solar panels,” Energies, vol. 15, no. 22, Art. no. 8100, 2022.
• 4. H. Malik, M. Alsabban, and S. M. Qaisar, “Arduino based automatic solar panel dust disposition estimation and cloud based reporting,” Procedia Comput. Sci., vol. 194, pp. 102–113, 2021.
• 5. K. A. Abuqaaud, and A. Ferrah, “A novel technique for detecting and monitoring dust and soil on solar photovoltaic panel,” in Proc. 2020 Advances Sci. Eng. Technol. Int. Conf. (ASET), Dubai, UAE, Feb, pp. 1–6.
• 6. D. Saquib, M. N. Nasser, and S. Ramaswamy, “Image processing based dust detection and prediction of power using ANN in PV systems,” in Proc. 2020 Third Int. Conf. Smart Syst. Inventive Technol. (ICSSIT), Tirunelveli, India, Aug, pp. 1286–1292.
• 7. S. Keerthana, S. Hariharasudhan, and M. Sundaram, “Image processingbased dust detection for solar panels,” in Proc. 2024 Int. Conf. Smart Syst. Electr. Electron. Commun. Comput. Eng. (ICSSEECC), Coimbatore, India, Jun. pp. 511–515, 2024.
• 8. J. Abukhait, “Dust detection on solar panels: A computer vision approach,” Ing. Syst. Inf., vol. 29, no. 2, pp. 533–541, 2024.
• 9. V. Kavya, and R. M. R. Keshav, “Solar dust detection system,” in Proc. 2018 Int. Conf. Power Energy, Environ. Intell. Control (PEEIC), Greater Noida, India, Apr. New York: IEEE, 2018, pp. 138–140.
• 10. A. M. Alatwi, H. Albalawi, A. Wadood, H. Anwar, and H. M. El-Hageen, “Deep learning-based dust detection on solar panels: A low-cost sustainable solution for increased solar power generation,” Sustainability, vol. 16, no. 19, Art. no. 8664, 2024.
• 11. A. Mobiny, A. Singh, and H. V. Nguyen, “Risk-aware machine learning classifier for skin lesion diagnosis,” J. Clin. Med., vol. 8, no. 8, Art. no. 1241, 2019.
• 12. K. Nair, A. Deshpande, R. Guntuka, and A. Patil, “Analysing X-ray images to detect lung diseases using DenseNet-169 technique,” SSRN Electron. J., 2022.
• 13. R. T. Wahyuningrum, A. Kusumaningsih, W. Putra Rajeb, and I. K. E. Purnama, “Classification of corn leaf disease using an optimized DenseNet-169 model,” in Proc. 2021 9th Int. Conf. Big Data Inf., Technol. (ICBIT), Surabaya, Indonesia, Dec, 2021, pp. 67–73.
• 14. A. Mehmood, “Efficient anomaly detection in crowd videos using pretrained 2D convolutional neural networks,” IEEE Access, vol. 9, pp. 138283–138295, 2021.
• 15. S. N. Endah, and I. N. Shiddiq, “Transfer learning of Xception architecture for garbage classification,” in Proc. 2020 4th Int. Conf. Informatics Comput. Sci. (ICICoS), Yogyakarta, Indonesia, Nov., 2020, pp. 1–4.
• 16. S. H. Kassani, P. H. Kassani, R. Khazaeinezhad, M. J. Wesolowski, K. A. Schneider, and R. Deters, “Diabetic retinopathy classification using a modified Xception architecture,” in Proc. 2019 IEEE Int. Symp. Signal Process. Inf. Technol. (ISSPIT), Atlanta, GA, USA: Dec, 2019, pp. 1–6.
• 17. L. Ali, F. Alnajjar, H. A. Jassmi, M. Gocho, W. Khan, and M. A. Serhani, “Performance evaluation of deep CNN-based crack detection and localization techniques for concrete structures,” Sensors (Basel), vol. 21, no. 5, Art. no. 1688, 2021.
• 18. X. Xia, C. Xu, and B. Nan, “Inception-v3 for flower classification,” in Proc. 2017 2nd Int. Conf. Image, Vision Comput. (ICIVC), Shenzhen, China, Jun, 2017, pp. 783–787.
• 19. S. M. Sam, K. Kamardin, N. N. A. Sjarif, and N. Mohamed, “Offline signature verification using deep learning convolutional neural network architectures GoogLeNet Inception-v1 and Inception-v3,” Procedia Comput. Sci., vol. 161, pp. 475–483, 2019.
• 20. R. Polikar, “Ensemble learning,” in Ensemble Machine Learning: Methods and Applications, C. Zhang, Ed. Cham, Switzerland: Springer, 2012, pp. 1–34.
• 21. H. Lappalainen, and J. W. Miskin, “Ensemble learning,” in Advances in Independent Component Analysis, M. Girolami, Ed. London, UK: Springer, 2000, pp. 75–92.
• 22. A. Dogan, and D. Birant, “A weighted majority voting ensemble approach for classification,” in Proc. 2019 4th Int. Conf. Comput. Sci. Eng. (UBMK), Elazığ, Turkey, Sep, 2019, pp. 1–6.
• 23. H. Wang, Y. Yang, H. Wang, and D. Chen, “Soft-voting clustering ensemble,” in Multiple Classifier Systems: 11th Int. Workshop, MCS 2013, Proc., Nanjing, China, May, pp. 307–318, 2013.
• 24. M. A. Khan, N. Iqbal, H. Jamil, H. Jamil, and D. H. Kim, “An optimized ensemble prediction model using AutoML based on soft voting classifier for network intrusion detection,” J. Netw. Comput. Appl., vol. 212, Art. no. 103560, 2023.
• 25. K. Tao et al., “Unlocking potential of pyrochlore in energy systems via soft voting ensemble learning,” Small, vol. 20, no. 42, Art. no. 2402756, 2024.
• 26. M. Ertel, A. Sadquı, S. Amalı, I. Mahmoudı, Y. Bouferma, and N. E. El Faddoulı, “Predicting the severity of new SARS-CoV-2 variants in vaccinated patients using machine learning,” J. Theor. Appl. Inf. Technol., vol. 101, no. 10, p. 2023, 2023.
• 27. R. Yacouby, and D. Axman, “Probabilistic extension of precision, recall, and F1-score for a more comprehensive evaluation of classification models,” in Proc. 2020 First Workshop Eval. Comparison NLP Syst. (EvalNLPSys), Helsinki, Finland, Nov., 2020, pp. 79–91.
• 28. Y. Liu, Y. Zhou, S. Wen, and C. Tang, “A strategy on selecting performance metrics for classifier evaluation,” Int. J. Mob. Comput. Multimedia Commun., vol. 6, no. 4, pp. 20–35, 2014.
• 29. V. S. S. Reddy, A. Sreejagathi, S. M. Venthan, and P. Vijayakumar, “Deep learning in solar panel maintenance: A model comparison for automated cleaning,” in Proc. 2025 Fifth Int. Conf. Adv. Electr. Comput. Commun. Sustain. Technol. (ICAECT), Istanbul, Turkey, Jan. pp. 1–7, 2025, pp. 1–7.