Performance Analysis of YOLO11 for Welding Defect Detection Under Low-Light Conditions
Abstract
This study aims to analyze the impact of image enhancement techniques on welding defect detection performance using a deep learning-based YOLO11L model. The dataset consists of 1392 welding images categorized into four classes: Good, Crack, Porosity, and Bad, with a significant class imbalance. Five image enhancement methods were evaluated, namely Zero-DCE, RETINEX, CLAHE, Supervision, and Gamma Correction, and compared against a no-enhancement baseline. Image quality was assessed using SSIM, and PSNR, while detection performance was evaluated using Precision, Recall, F1-Score, and mAP50. The results show that Gamma Correction achieves the best image quality improvement, with an average SSIM of 0.569, and a PSNR of 18.862 dB. However, contrasting results are observed at the detection stage, where 0.7772 and 0.6969, respectively, for the Gamma Correction-based model while for the baseline model without enhancement outperforms the enhanced model, achieving a mAP50 of 0.7098 and an F1-Score of 0.6965. This finding reveals a paradox where improved visual image quality does not necessarily lead to better object detection performance. This study highlights the importance of end-to-end evaluation in computer vision systems, particularly in industrial inspection applications, and demonstrates that original images, which are closer to the pretrained data distribution, may yield better detection results than heavily enhanced images.
Keywords
Full Text:
PDFReferences
Y. Pernando, F. L. Gaol, H. Soeparno, and Y. Arifin, “Software Quality Assessment Methods and Standards in Weld Defect Detection for Shipbuilding,” 2024 6th International Conference on Cybernetics and Intelligent System, ICORIS 2024, pp. 1–6, 2024, doi: 10.1109/ICORIS63540.2024.10903940.
Y. Pernando, N. A. Catur Andryani, A. Chowanda, and W. Budiharto, “Low-Light Welding Defect Detection Using Retinex-Enhanced YOLOv11 for Steel Welding Inspection,” 2025 IEEE International Conference on Communication, Networks and Satellite (COMNETSAT), no. December, pp. 77–83, 2026, doi: 10.1109/comnetsat68601.2025.11324576.
A. A. Melakhsou and M. Batton-Hubert, “On welding defect detection and causalities between welding signals,” 2021 IEEE 17th International …, 2021, [Online]. Available: https://ieeexplore.ieee.org/abstract/document/9551659/
Y. Gao, P. Zhong, X. Tang, H. Hu, and P. Xu, “Feature extraction of laser welding pool image and application in welding quality identification,” IEEE Access, 2021, [Online]. Available: https://ieeexplore.ieee.org/abstract/document/9524618/
J. E. Kwon, J. H. Park, J. H. Kim, Y. H. Lee, and S. I. Cho, “Context and scale-aware YOLO for welding defect detection,” NDT &E International, 2023, [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0963869523001342
Q. Tu, H. Liu, C. Qu, L. Tian, and D. Zhu, “An accurate detection method for randomly distributed welding slags using an improved Yolo v3 network,” International Journal of …, 2021, doi: 10.1504/IJCMSSE.2021.121357.
W. Guo, L. Huang, and L. Liang, “A weld seam dataset and automatic detection of welding defects using convolutional neural network,” The 8th International Conference on Computer …, 2020, doi: 10.1007/978-3-030-14680-1_48.
L. Liang, “A Weld Seam Dataset and Automatic Detection of Welding Defects Using Convolutional Neural Network,” The 8th International Conference on Computer …, 2019, [Online]. Available: https://books.google.com/books?hl=en&lr=&id=Cp6SDwAAQBAJ&oi=fnd&pg=PA434&dq=welding+defect+cnn&ots=q06tWIRyks&sig=WWHH8SQp4TMJZ9X53BGojGtAXN4
R. K. Sharma, A. D. Pakki, and J. Holovský, “Silicon heterojunction solar cells: Excellent candidate for low light illuminations,” Solar Energy Materials and Solar Cells, 2024, [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0927024824003131
J. He, “A Zero-Dce-Based Algorithm for Enhancing Low-Light Images with Uneven Illumination,” Available at SSRN 4837890, 2024, [Online]. Available: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4837890
C. Shi et al., “RICNET: Retinex-Inspired Illumination Curve Estimation for Low-Light Enhancement in Industrial Welding Scenes,” Sensors, vol. 25, no. 16, pp. 1–19, 2025, doi: 10.3390/s25165192.
U. Kumaran, E. R. Ashwin, and ..., “Web application for real-time enhancement of low-light images using Zero-DCE and hybrid optimization,” 2025 3rd International …, 2025, [Online]. Available: https://ieeexplore.ieee.org/abstract/document/10969162/
Z. Chen, J. Yang, F. Li, and Z. Feng, “Real-Time Low-Light Image Enhancement Method for Train Driving Scene Based on Improved Zero-DCE,” International Conference on Electrical and …, 2023, doi: 10.1007/978-981-99-9319-2_2.
S. S. M. Sheet, T. S. Tan, M. A. As’ari, W. H. W. Hitam, and J. S. Y. Sia, “Retinal disease identification using upgraded CLAHE filter and transfer convolution neural network,” ICT Express, vol. 8, no. 1, pp. 142–150, Mar. 2022, doi: 10.1016/j.icte.2021.05.002.
J. Wang, H. Wang, Y. Sun, and J. Yang, “Improved retinex-theory-based low-light image enhancement algorithm,” 2023, mdpi.com. [Online]. Available: https://www.mdpi.com/2076-3417/13/14/8148
H. Zhao, S. Xu, L. Peng, H. Hu, and S. Jiang, “Efficient Gamma-Based Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement,” 2025, mdpi.com. [Online]. Available: https://www.mdpi.com/2076-3417/15/13/7382
W. Dai, D. Li, D. Tang, H. Wang, and Y. Peng, “Deep learning approach for defective spot welds classification using small and class-imbalanced datasets,” Neurocomputing, 2022, [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0925231222000042
R. Yulianto, M. S. Rusli, A. S. B. Karno, W. Hastomo, and ..., Innovative UNET-Based Steel Defect Detection Using 5 Pretrained Models. catalog.lib.kyushu-u.ac.jp, 2023. [Online]. Available: https://catalog.lib.kyushu-u.ac.jp/ja/recordID/7160923/?repository=yes
R. Yulianto, M. S. Rusli, A. S. B. Karno, W. Hastomo, and ..., Innovative UNET-Based Steel Defect Detection Using 5 Pretrained Models. catalog.lib.kyushu-u.ac.jp, 2023. [Online]. Available: https://catalog.lib.kyushu-u.ac.jp/ja/recordID/7160923/?repository=yes
S. A. Rizvi and W. Alib, “Welding defects, Causes and their Remedies: A Review,” Teknomekanik, 2019, [Online]. Available: http://teknomekanik.ppj.unp.ac.id/index.php/teknomekanik/article/view/32
B. Zhang, K. M. Hong, and Y. C. Shin, “Deep-learning-based porosity monitoring of laser welding process,” Manuf. Lett., vol. 23, pp. 62–66, Jan. 2020, doi: 10.1016/j.mfglet.2020.01.001.
DOI: https://doi.org/10.55311/aiocsit.v7i1.370
Refbacks
- There are currently no refbacks.