Journal of sustainable Energy Systems

Journal of sustainable Energy Systems

Intelligent Detection of Surface Defects in Solar Panels Using Deep Learning-Based Computer Vision

Document Type : Original Article

Author
Ali Jelokhani
School of Electrical and Computer Engineering, University of Tehran, Tehran, Iran
10.22059/ses.2026.414893.1244
Abstract
With the increasing use of photovoltaic systems, monitoring the health of solar panels has become an important issue for sustainable operation, reducing maintenance costs, and improving energy generation efficiency. Defects such as breakage, dust, shadow, and surface variations can reduce panel performance and shorten its service life. Traditional inspection methods are generally time-consuming, dependent on human labor, and inefficient on a large scale. In this study, a computer vision and deep learning-based framework is proposed for detecting and localizing surface regions in solar panel images. The proposed model consists of a ResNet101 feature extractor, a Sobel operator-based edge extraction module, a spatial–channel attention module, an FPN/PAN multi-scale feature fusion structure, and a multi-scale detection head. The model output includes bounding boxes, region classes, and confidence scores; therefore, it can simultaneously identify multiple different regions in a single image. The dataset used in this study was constructed by combining images collected from public sources with manually collected data and includes four classes: breakage, clean region, dust, and shadow. The experimental results showed that the proposed model, achieving a Precision of 0.954, Recall of 0.939, F1-score of 0.946, mAP@0.5 of 0.966, and mAP@0.5:0.95 of 0.766, outperformed baseline models and recent photovoltaic defect detection methods. Furthermore, the ablation study results showed that combining edge information, spatial–channel attention, and multi-scale feature fusion plays an important role in improving detection and localization accuracy. Based on the results, the proposed method can be used as a practical solution for automatic monitoring, targeted cleaning, and predictive maintenance of solar panels.
Keywords
Solar panel
Solar panel defect detection
Computer vision
Deep learning
Subjects
[1] W. Tang, Q. Yang, Z. Dai, and W. Yan, "Module defect detection and diagnosis for intelligent maintenance of solar photovoltaic plants: Techniques, systems and perspectives," Energy, vol. 297, p. 131222, 2024/06/15/ 2024, doi: https://doi.org/10.1016/j.energy.2024.131222.
[2] A. Thakfan and Y. Bin Salamah, "Artificial-Intelligence-Based Detection of Defects and Faults in Photovoltaic Systems: A Survey," Energies, vol. 17, no. 19, p. 4807, 2024. [Online]. Available: https://www.mdpi.com/1996-1073/17/19/4807.
[3] M. Jalal, I. U. Khalil, and A. u. Haq, "Deep learning approaches for visual faults diagnosis of photovoltaic systems: State-of-the-Art review," Results in Engineering, vol. 23, p. 102622, 2024/09/01/ 2024, doi: https://doi.org/10.1016/j.rineng.2024.102622.
[4] D. Lang and Z. Lv, "A PV cell defect detector combined with transformer and attention mechanism," Scientific Reports, vol. 14, no. 1, p. 20671, 2024/09/05 2024, doi: 10.1038/s41598-024-72019-5.
[5] W. Cai and H. Lv, "YOLO-PV: An Enhanced YOLO11n Model with Multi-Scale Feature Fusion for Photovoltaic Panel Defect Detection," Energies, vol. 18, no. 20, p. 5489, 2025. [Online]. Available: https://www.mdpi.com/1996-1073/18/20/5489.
[6] A. Mohamed et al., "Optimized YOLO based model for photovoltaic defect detection in electroluminescence images," Scientific Reports, vol. 15, no. 1, p. 32955, 2025/09/26 2025, doi: 10.1038/s41598-025-13956-7.
[7] A. Rahman, "Solar Panel Surface Defect and Dust Detection: Deep Learning Approach," Journal of Imaging, vol. 11, no. 9, p. 287, 2025. [Online]. Available: https://www.mdpi.com/2313-433X/11/9/287.
[8] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 27-30 June 2016 2016, pp. 779-788, doi: 10.1109/CVPR.2016.91.
[9] K. He, X. Zhang, S. Ren, and J. Sun, "Deep Residual Learning for Image Recognition," in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 27-30 June 2016 2016, pp. 770-778, doi: 10.1109/CVPR.2016.90.
[10]           A. Guha Roy, N. Navab, and C. Wachinger, "Concurrent Spatial and Channel Squeeze & Excitation in Fully Convolutional Networks," 03/07 2018, doi: 10.48550/arXiv.1803.02579.
[11]           S. Deitsch et al., "Automatic classification of defective photovoltaic module cells in electroluminescence images," Solar Energy, vol. 185, pp. 455-468, 2019/06/01/ 2019, doi: https://doi.org/10.1016/j.solener.2019.02.067.
[12]           W. Tang, Q. Yang, K. Xiong, and W. Yan, "Deep learning based automatic defect identification of photovoltaic module using electroluminescence images," Solar Energy, vol. 201, pp. 453-460, 2020/05/01/ 2020, doi: https://doi.org/10.1016/j.solener.2020.03.049.
[13]           M. W. Akram, G. Li, Y. Jin, X. Chen, C. Zhu, and A. Ahmad, "Automatic detection of photovoltaic module defects in infrared images with isolated and develop-model transfer deep learning," Solar Energy, vol. 198, pp. 175-186, 2020/03/01/ 2020, doi: https://doi.org/10.1016/j.solener.2020.01.055.
[14]           S. Ren, K. He, R. Girshick, and J. Sun, "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 39, no. 6, pp. 1137-1149, 2017, doi: 10.1109/TPAMI.2016.2577031.
[15]           W. Liu et al., "SSD: Single Shot MultiBox Detector," in Computer Vision – ECCV 2016, Cham, B. Leibe, J. Matas, N. Sebe, and M. Welling, Eds., 2016// 2016: Springer International Publishing, pp. 21-37.
[16]           T. Y. Lin, P. Goyal, R. Girshick, K. He, and P. Dollár, "Focal Loss for Dense Object Detection," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 42, no. 2, pp. 318-327, 2020, doi: 10.1109/TPAMI.2018.2858826.
[17]           N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, "End-to-End Object Detection with Transformers," in Computer Vision – ECCV 2020, Cham, A. Vedaldi, H. Bischof, T. Brox, and J.-M. Frahm, Eds., 2020// 2020: Springer International Publishing, pp. 213-229.
[18]           W. Pan, X. Sun, Y. Wang, Y. Cao, Y. Lang, and Y. Qian, "Enhanced photovoltaic panel defect detection via adaptive complementary fusion in YOLO-ACF," Scientific Reports, vol. 14, no. 1, p. 26425, 2024/11/02 2024, doi: 10.1038/s41598-024-75772-9.
[19]           Y. Jia, G. Chen, and L. Zhao, "Defect detection of photovoltaic modules based on improved VarifocalNet," Scientific Reports, vol. 14, no. 1, p. 15170, 2024/07/02 2024, doi: 10.1038/s41598-024-66234-3.
[20]           Z. Qu et al., "A photovoltaic cell defect detection model capable of topological knowledge extraction," Scientific Reports, vol. 14, no. 1, p. 21904, 2024/09/19 2024, doi: 10.1038/s41598-024-72717-0.
[21]           N. Wang, S. Huang, X. Liu, Z. Wang, Y. Liu, and Z. Gao, "MRA-YOLOv8: A Network Enhancing Feature Extraction Ability for Photovoltaic Cell Defects," Sensors, vol. 25, no. 5, p. 1542, 2025. [Online]. Available: https://www.mdpi.com/1424-8220/25/5/1542.
[22]           H. Xue, L. Liu, Q. Wu, J. He, and Y. Fan, "Defect Detection Algorithm for Photovoltaic Cells Based on SEC-YOLOv8," Processes, vol. 13, no. 8, p. 2425, 2025. [Online]. Available: https://www.mdpi.com/2227-9717/13/8/2425.
[23]           C. Chen, Z. Chen, H. Li, Y. Wang, G. Lei, and L. Wu, "Research on Defect Detection in Lightweight Photovoltaic Cells Using YOLOv8-FSD," Sensors, vol. 25, no. 3, p. 843, 2025. [Online]. Available: https://www.mdpi.com/1424-8220/25/3/843.
[24]           Y. Lu, C. Du, X. Li, S. Liang, Q. Zhang, and Z. Zhao, "A High-Precision Defect Detection Approach Based on BiFDRep-YOLOv8n for Small Target Defects in Photovoltaic Modules," Energies, vol. 18, no. 9, p. 2299, 2025. [Online]. Available: https://www.mdpi.com/1996-1073/18/9/2299.
[25]           M. Al-Mahbashi et al., "An effective approach to improving photovoltaic defect detection using the new DCD-YOLOv8s model," Scientific Reports, vol. 15, no. 1, p. 38308, 2025/11/03 2025, doi: 10.1038/s41598-025-22307-5.
[26]           W. Jiang and W. Liu, "Study on defect detection and identification of solar photovoltaic (PV) panels based on improved YOLO11," Scientific Reports, 2026/04/24 2026, doi: 10.1038/s41598-026-48824-5.
[27]           S. Meng and T. Xu, "Inspection Method Enabled by Lightweight Self-Attention for Multi-Fault Detection in Photovoltaic Modules," Electronics, 2025.
[28]           M. Dhimish, "HOTSPOT-YOLO: A Lightweight Deep Learning Attention-Driven Model for Detecting Thermal Anomalies in Drone-Based Solar Photovoltaic Inspections," ArXiv, vol. abs/2508.18912, 2025.
[29]           H. He et al., "A Self-Evolving Defect Detection Framework for Industrial Photovoltaic Systems," arXiv preprint arXiv:2603.14869, 2026.
[30]           D. Grzechca, F. Ez-Zahiri, Ł. Chruszczyk, and F. Bian, "Defect-Aware RGB Representation and Resolution-Efficient Deep Learning for Photovoltaic Failure Detection in Electroluminescence Images," Applied Sciences, vol. 16, no. 4, p. 2148, 2026. [Online]. Available: https://www.mdpi.com/2076-3417/16/4/2148.
[31]           F. M. Talaat, M. Salem, and W. M. Shaban, "AI-driven fault detection and classification in photovoltaic systems using deep learning techniques," Scientific Reports, vol. 16, no. 1, p. 8727, 2026/03/10 2026, doi: 10.1038/s41598-026-40246-7.
[32]           A. Seghiour et al., "Fault detection and diagnosis in photovoltaic systems using artificial intelligence and time–frequency analysis," Scientific Reports, vol. 16, no. 1, p. 10056, 2026/02/20 2026, doi: 10.1038/s41598-026-39386-7.
[33]           S. R. Mohanty, M. U. Maruf, V. Singh, and Z. Ahmad, "Machine learning approaches for automatic defect detection in photovoltaic systems," Solar Energy, vol. 298, p. 113672, 2025/09/15/ 2025, doi: https://doi.org/10.1016/j.solener.2025.113672.
[34]           E. Zaman and R. Khanam, PV-faultNet: Optimized CNN Architecture to detect defects resulting efficient PV production. 2024.
[35]           J. Zhang, X. Chen, H. Wei, and K. Zhang, "A lightweight network for photovoltaic cell defect detection in electroluminescence images based on neural architecture search and knowledge distillation," Applied Energy, vol. 355, p. 122184, 2024/02/01/ 2024, doi: https://doi.org/10.1016/j.apenergy.2023.122184.
 
Journal of sustainable Energy Systems
Volume 5, Issue 4
Autumn 2026
Pages 665-689