Vision Transformer-Based Visual Inspection of Photovoltaic Cells from Angular Images and Fault Recognition
Hiren MEWADA, Syam Sundar LINGALA, Nayeemuddin MOHAMMED
Abstract. The most predictable and economical renewable energy source is solar photovoltaic (PV). Nevertheless, various environmental factors cause physical and electrical defects in the PV cell, reducing its power output, reliability, and longevity. This paper addresses this challenge using a vision transformer (VT)-based CNN. A vision transformer transforms an image into patches, whereas a deep CNN does not. Long-range associations are absent in a deep CNN. The photovoltaic cells exhibit complex, overlapping defects in a specific area. Hence, patch-based feature extraction on the VT result yields higher detection rates. The proposed model is tested to assess the extent of electrical and physical damage in cells relative to healthy cells. The data includes pictures taken in different angles and different environmental conditions, and the information obtained was made consistent, reliable, and applicable in the field. The proposed model was effective in classifying PV cells, achieving a precision of 0.9722 for clean images, 0.9565 for clean images, and 0.9286 for electrical- and physical-damage images, respectively.
Keywords
Fault Recognition, CNN, Renewable Energy, Photovoltaic, Vision Transformer
Published online 4/25/2026, 9 pages
Copyright © 2026 by the author(s)
Published under license by Materials Research Forum LLC., Millersville PA, USA
Citation: Hiren MEWADA, Syam Sundar LINGALA, Nayeemuddin MOHAMMED, Vision Transformer-Based Visual Inspection of Photovoltaic Cells from Angular Images and Fault Recognition, Materials Research Proceedings, Vol. 64, pp 21-29, 2026
DOI: https://doi.org/10.21741/9781644904091-3
The article was published as article 3 of the book Energy Futures
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 license. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
References
[1] Z. Dong, Q. Zhang, X. Shu, et al. Photo-assisted charging of heterostructured NiCo2S4@NiCo-LDH composite electrode with remarkable photoelectronic memory effect for high-performance asymmetric supercapacitor, Energy Convers Management, 315 (2024), 118769 https://doi.org/10.1016/j.enconman.2024.118769
[2] C. Zhang, L. Tu, Z. Yang, et al. A CMMOG-based lithium-battery SOH estimation method using multi-task learning framework, J Energy Storage, 107 (2025), 114884 https://doi.org/10.1016/j.est.2024.114884
[3] Z. Wu, Y. Hu, J.X. Wen, et al. A review for solar panel fire accident prevention in large-scale PV applications, IEEE Access, 8 (2020), pp. 132466-132480 https://doi.org/10.1109/ACCESS.2020.3010212
[4] O. Idbouhouch, N. Rabbah, N. Lamrini, et al. Assessing PV inverter efficiency degradation under semi-arid conditions: a case study in Morocco, Heliyon, 10 (17) (2024), e36906 https://doi.org/10.1016/j.heliyon.2024.e36906
[5] M. Jalal, I.U. Khalil, A.U. Haq, Deep learning approaches for visual faults diagnosis of photovoltaic systems: state-of-the-Art review, Results Eng, 23 (2024), 102622. https://doi.org/10.1016/j.rineng.2024.102622
[6] Z. Dong, M. Chen, D. Qin, et al. Recent advances and perspective of modified TiO2-based photoanodes toward photoelectrochemical water splitting, Fuel, 373 (2024), 132366 https://doi.org/10.1016/j.fuel.2024.132366
[7] B. Taghezouit, F. Harrou, Y. Sun, et al. Model-based fault detection in photovoltaic systems: a comprehensive review and avenues for enhancement, Results Eng, 21 (2024), 101835. https://doi.org/10.1016/j.rineng.2024.101835
[8] Z. Fu, X. Shu, Q. Zhang, et al. Solar-driven induced photoelectron remember effect involved in core-shell NiCo2S4@Ni3V2O8 composite electrode with superior electrochemical energy storage for asymmetric supercapacitor, Energy Convers Manag, 323 (2025), 119190 https://doi.org/10.1016/j.enconman.2024.119190
[9] Z. Dong, D. Qin, J. Ma, et al. Bulk-phase and surface dual-defective engineering enabled Ti-based nanotubes photoanode for highly efficient photoelectrochemical water splitting, Renew Energy, 231 (2024) 121004. https://doi.org/10.1016/j.renene.2024.121004
[10] A.F. Amiri, H. Oudira, A. Chouder, et al., Faults detection and diagnosis of PV systems based on machine learning approach using random forest classifier, Energy Convers Manag, 301 (2024), 118076. https://doi.org/10.1016/j.enconman.2024.118076
[11] M. Simon, E.L. Meyer, Detection and analysis of hot-spot formation in solar cells, Sol Energy Mater Sol Cell, 94 (2) (2010), pp. 106-113. https://doi.org/10.1016/j.solmat.2009.09.016
[12] A. Mellit, An embedded solution for fault detection and diagnosis of photovoltaic modules using thermographic images and deep convolutional neural networks, Eng Appl Artif Intell, 116 (2022), 105459. https://doi.org/10.1016/j.engappai.2022.105459
[13] V. Kirubakaran, D.M.D. Preethi, U. Arunachalam, et al., Infrared thermal images of solar PV panels for fault identification using image processing technique, Int J Photoenergy, 2022 (2022), pp. 1-10 https://doi.org/10.1155/2022/6427076
[14] M. Simon, E.L. Meyer, Detection and analysis of hot-spot formation in solar cells, Sol Energy Mater Sol Cell, 94 (2) (2010), pp. 106-113. https://doi.org/10.1016/j.solmat.2009.09.016
[15] S. Boubaker, S. Kamel, N. Ghazouani, et al., Assessment of machine and deep learning approaches for fault diagnosis in photovoltaic systems using infrared thermography, Remote Sens, 15 (6) (2023), p. 1686 https://doi.org/10.3390/rs15061686
[16] K. Masita, A. Hasan, T. Shongwe, et al. Deep learning in defects detection of PV modules: a review, Solar Energy Advances, 5 (2025), 100090. https://doi.org/10.1016/j.seja.2025.100090
[17] A.W. Kandeal, M.R. Elkadeem, A. Kumar Thakur, et al., Infrared thermography-based condition monitoring of solar photovoltaic systems: a mini review of recent advances, Sol Energy, 223 (2021), pp. 33-43 https://doi.org/10.1016/j.solener.2021.05.032
[18] C. Bu, T. Liu, R. Li, et al., Electrical pulsed infrared thermography and supervised learning for PV cells defects detection, Sol Energy Mater Sol Cell, 237 (2022), 111561. https://doi.org/10.1016/j.solmat.2021.111561
[19] B. Li, C. Delpha, D. Diallo, et al. Application of artificial neural networks to photovoltaic fault detection and diagnosis: a review, Renew Sustain Energy Rev, 138 (2021), 110512. https://doi.org/10.1016/j.rser.2020.110512
[20] N.C. Yang, M. FaizanLong short-term memory-based feedforward neural network algorithm for photovoltaic fault detection under irradiance conditions, IEEE Trans Instrum Meas, 73 (2024), pp. 1-11. https://doi.org/10.1109/TIM.2024.3413159
[21] A. Hichri, M. Hajji, M. Mansouri, et al., Genetic-algorithm-based neural network for fault detection and diagnosis: application to grid-connected photovoltaic systems, Sustainability, 14 (17) (2022), 10518. https://doi.org/10.3390/su141710518
[22] S. Han, Y. Yu, Y. Guo, Fault diagnosis of open-circuit in photovoltaic inverters based on improved GAF-SE-ResNet, Journal of Solar Energy, 45 (10) (2024), pp. 336-344
[23] Ledmaoui, Y. Solar-Panel-Anomalies-Detecting. GitHub Repository. 2025. Available online: https://github.com/Ledmaoui/Solar-Panel-Anomalies-Detecting (accessed on December 2025).
[24] Alanazi, M.; Alamri, Y.A. SolarFaultAttentionNet: Dual-Attention Framework for Enhanced Photovoltaic Fault Classification. Inventions 2025, 10, 91. https://doi.org/10.3390/inventions10050091
“”
