基于SPDConv-EMA与自适应加权融合的改进YOLO11n风机叶片缺陷检测方法
An Improved YOLO11n Method for Wind Turbine Blade Defect Detection Based on SPDConv-EMA and Adaptive Weighted Fusion YOLO11n
DOI: 10.12677/csa.2026.168263, PDF,    科研立项经费支持
作者: 朱 向, 张伶俐：重庆文理学院群与图的理论及应用重庆市高校重点实验室，重庆；卢成武*：重庆文理学院群与图的理论及应用重庆市高校重点实验室，重庆；重庆第二师范学院数学与大数据学院，重庆
关键词: 风机叶片；表面缺陷检测；YOLO11n；SPDConv；C3k2_EMA；AWFF；Wind Turbine Blade； Surface Defect Detection； YOLO11n； SPDConv； C3k2_EMA； AWFF
摘要: 针对无人机航拍风机叶片缺陷检测中小目标细节易丢失、缺陷边缘纹理特征不明显、复杂背景干扰明显及多尺度特征融合不足等问题，本文提出一种融合SPDConv、C3k2_EMA与AWFF的改进YOLO11n检测方法。该方法在Backbone关键下采样位置采用SPDConv替换普通跨步卷积，以减少裂纹、边缘磨损等细长小目标在降采样过程中的空间信息损失；在深层特征提取单元中嵌入EMA注意力机制，构建C3k2_EMA模块，以增强缺陷区域响应并抑制阴影、反光和叶片纹理等背景干扰；在Neck关键融合节点引入AWFF模块，通过可学习权重自适应平衡浅层细节与深层语义特征。实验结果表明，改进模型的Precision、Recall、mAP@0.5和mAP@0.5:0.95分别达到84.8%、84.3%、86.7%和57.6%，较原始YOLO11n分别提高1.2、3.9、2.5和2.4个百分点，验证了本文方法在风机叶片表面缺陷检测任务中的有效性。
Abstract: To address the loss of small-defect details, weak edge-texture representation, complex background interference, and insufficient multi-scale feature fusion in UAV-based wind turbine blade inspection, this paper proposes an improved YOLO11n method integrating SPDConv, C3k2_EMA, and AWFF. SPDConv is used to replace ordinary strided convolutions at key down sampling positions in the backbone, reducing spatial information loss for slender cracks and edge defects. The EMA attention mechanism is embedded into deep feature extraction units to construct C3k2_EMA, which enhances responses to defective regions while suppressing shadows, highlights, and blade texture interference. In the neck, AWFF is introduced at key fusion nodes to adaptively balance shallow detail features and deep semantic features through learnable weights. Experimental results show that the improved model achieves Precision, Recall, mAP@0.5, and mAP@0.5:0.95 of 84.8%, 84.3%, 86.7%, and 57.6%, respectively, improving the original YOLO11n by 1.2, 3.9, 2.5, and 2.4 percentage points and demonstrating its effectiveness for wind turbine blade surface defect detection.
文章引用：朱向, 张伶俐, 卢成武. 基于SPDConv-EMA与自适应加权融合的改进YOLO11n风机叶片缺陷检测方法[J]. 计算机科学与应用, 2026, 16(8): 65-79. https://doi.org/10.12677/csa.2026.168263

参考文献

[1] 刘俊伶.《风能北京宣言2.0》发布未来五年装机目标翻倍式上调[N]. 证券时报, 2025-10-21(A06).
[2] Shihavuddin, A.S.M., Chen, X., Fedorov, V., Nymark Christensen, A., Andre Brogaard Riis, N., Branner, K., et al. (2019) Wind Turbine Surface Damage Detection by Deep Learning Aided Drone Inspection Analysis. Energies, 12, Article 676.
https://doi.org/10.3390/en12040676
[3] Xu, D., Wen, C. and Liu, J. (2019) Wind Turbine Blade Surface Inspection Based on Deep Learning and UAV-Taken Images. Journal of Renewable and Sustainable Energy, 11, Article No. 053305.
https://doi.org/10.1063/1.5113532
[4] Yang, X., Zhang, Y., Lv, W. and Wang, D. (2021) Image Recognition of Wind Turbine Blade Damage Based on a Deep Learning Model with Transfer Learning and an Ensemble Learning Classifier. Renewable Energy, 163, 386-397.
https://doi.org/10.1016/j.renene.2020.08.125
[5] Heo, S.J. and Na, W.S. (2025) Review of Drone-Based Technologies for Wind Turbine Blade Inspection. Electronics, 14, Article 227.
https://doi.org/10.3390/electronics14020227
[6] Shihavuddin, A.S.M., Rashid, M.R.A., Maruf, M.H., Hasan, M.A., Haq, M.A.U., Ashique, R.H., et al. (2021) Image Based Surface Damage Detection of Renewable Energy Installations Using a Unified Deep Learning Approach. Energy Reports, 7, 4566-4576.
https://doi.org/10.1016/j.egyr.2021.07.045
[7] Ren, S., He, K., Girshick, R., et al. (2015) Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. Advances in Neural Information Processing Systems 28, 2015, 91-99.
[8] Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C., et al. (2016) SSD: Single Shot Multibox Detector. In: Lecture Notes in Computer Science, Springer International Publishing, 21-37.
https://doi.org/10.1007/978-3-319-46448-0_2
[9] Redmon, J., Divvala, S., Girshick, R. and Farhadi, A. (2016) You Only Look Once: Unified, Real-Time Object Detection. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 27-30 June 2016, 779-788.
https://doi.org/10.1109/cvpr.2016.91
[10] Wang, C., Bochkovskiy, A. and Liao, H.M. (2023) YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art for Real-Time Object Detectors. 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, 17-24 June 2023, 7464-7475.
https://doi.org/10.1109/cvpr52729.2023.00721
[11] Wang, C., Yeh, I. and Mark Liao, H. (2024) YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information. In: Leonardis, A., Ricci, E., Roth, S., Russakovsky, O., Sattler, T. and Varol, G. Eds., Lecture Notes in Computer Science, Springer, 1-21.
https://doi.org/10.1007/978-3-031-72751-1_1
[12] Wang, A., Chen, H., Liu, L., Chen, K., Lin, Z., Han, J., et al. (2024) YOLOv10: Real-Time End-to-End Object Detection. Advances in Neural Information Processing Systems 37, Vancouver, 10-15 December 2024, 107984-108011.
https://doi.org/10.52202/079017-3429
[13] 王俊, 高贵兵. 一种改进YOLOv5s算法的风机叶片表面缺陷检测方法[J/OL]. 中国机械工程: 1-11.
https://link.cnki.net/urlid/42.1294.th.20250403.1149.002, 2026-06-05.
[14] 高文俊, 张海峰. 改进的YOLOv5风机叶片缺陷检测方法[J]. 建模与仿真, 2023, 12(4): 3574-3586.
[15] 曾勇杰, 范必双, 杨涯文, 等. 改进YOLOv8算法在风机叶片缺陷检测上的应用[J]. 电子测量与仪器学报, 2024, 38(8): 26-35.
[16] 李大华, 吴超强, 高强, 等. 改进YOLOv8n的风机桨叶表面缺陷轻量化检测网络[J]. 电子测量与仪器学报, 2025, 39(8): 145-155.
[17] 李松霖, 唐玲. 基于改进YOLO11的风机叶片表面缺陷检测研究[J/OL]. 电子测量技术: 1-17.
https://link.cnki.net/urlid/11.2175.TN.20260205.1612.011, 2026-06-05.
[18] Tong, K. and Wu, Y. (2022) Deep Learning-Based Detection from the Perspective of Small or Tiny Objects: A Survey. Image and Vision Computing, 123, Article 104471.
https://doi.org/10.1016/j.imavis.2022.104471
[19] Sunkara, R. and Luo, T. (2023) No More Strided Convolutions or Pooling: A New CNN Building Block for Low-Resolution Images and Small Objects. In: Amini, MR., Canu, S., Fischer, A., Guns, T., Kralj Novak, P. and Tsoumakas, G. Eds., Lecture Notes in Computer Science, Springer, 443-459.
https://doi.org/10.1007/978-3-031-26409-2_27
[20] Hu, J., Shen, L. and Sun, G. (2018) Squeeze-and-Excitation Networks. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 7132-7141.
https://doi.org/10.1109/cvpr.2018.00745
[21] Woo, S., Park, J., Lee, J. and Kweon, I.S. (2018) CBAM: Convolutional Block Attention Module. In: Ferrari, V., Hebert, M., Sminchisescu, C. and Weiss, Y., Eds., Lecture Notes in Computer Science, Springer International Publishing, 3-19.
https://doi.org/10.1007/978-3-030-01234-2_1
[22] Ouyang, D., He, S., Zhang, G., Luo, M., Guo, H., Zhan, J., et al. (2023) Efficient Multi-Scale Attention Module with Cross-Spatial Learning. ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Rhodes Island, 4-10 June 2023, 1-5.
https://doi.org/10.1109/icassp49357.2023.10096516
[23] Lin, T., Dollar, P., Girshick, R., He, K., Hariharan, B. and Belongie, S. (2017) Feature Pyramid Networks for Object Detection. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, 21-26 July 2017, 2117-2125.
https://doi.org/10.1109/cvpr.2017.106
[24] Liu, S., Qi, L., Qin, H., Shi, J. and Jia, J. (2018) Path Aggregation Network for Instance Segmentation. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 8759-8768.
https://doi.org/10.1109/cvpr.2018.00913
[25] Tan, M., Pang, R. and Le, Q.V. (2020) EfficientDet: Scalable and Efficient Object Detection. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, 13-19 June 2020, 10781-10790.
https://doi.org/10.1109/cvpr42600.2020.01079
[26] Liu, S., Huang, D. and Wang, Y. (2019) Learning Spatial Fusion for Single-Shot Object Detection. arXiv:1911.09516.