细粒度特征增强与尺寸匹配的光伏缺陷检测

doi:10.6040/j.issn.1672-3961.0.2024.170

摘要/Abstract

摘要： 针对光伏场站无人机巡检场景中,电池面板缺陷目标成像过小导致的漏检和误检问题,提出一种针对光伏电池板的高精度缺陷检测方法。根据无人机巡检中目标尺寸的分布特征,采用细粒度特征增强与尺寸匹配相结合的策略,提高光伏缺陷小目标检测精度。与传统样本数据增强、多尺度学习和特征增强等增强小目标策略不同,引入细节保留语义信息增强模块,在保留细粒度信息的同时,挖掘相关粗粒度语义细节。引入锚点-预测头匹配多尺度检测策略,确保锚点与特征图的尺寸匹配。本研究方法在PVEL-AD数据集上均值平均精度达到59.4%,在CARPK数据集上均值平均精度达到97.6%,对比主流目标检测模型,光伏缺陷目标检测性能显著提高。

关键词: 小目标检测, 实时目标检测, 缺陷检测, 特征增强, 尺寸匹配

Abstract: A high-precision defect detection method for photovoltaic panels was proposed to address the issue of missed and false detections associated with small defect targets in unmanned aerial vehicle(UAV)inspections of photovoltaic power stations. Considering the distribution characteristics of target sizes identified during UAV inspections, a strategy that combines fine-grained feature enhancement with scale matching was introduced to enhance the accuracy of small target defect detection in photovoltaics. Distinguished from traditional small target enhancement strategies including data augmentation, multi-scale learning, and feature enhancement, a detail-preserving semantic enhancement module was incorporated to retain fine-grained details and to mine related coarse-grained semantic details. A multi-scale detection strategy featuring anchor-prediction head matching was introduced to ensure the compatibility of anchor sizes with feature maps. The method achieved an average mean precision of 59.4% on the PVEL-AD dataset and 97.6% on the CARPK dataset, significantly improving the performance of photovoltaic defect target detection compared to mainstream object detection models.

Key words: small target detection, real-time object detection, defect detection, feature enhancement, scale matching

中图分类号:

TP391

索大翔,李波. 细粒度特征增强与尺寸匹配的光伏缺陷检测[J]. 山东大学学报 (工学版), 2025, 55(4): 9-17.

SUO Daxiang, LI Bo. Photovoltaic defect detection based on fine-grained feature enhancement and scale matching[J]. Journal of Shandong University(Engineering Science), 2025, 55(4): 9-17.

参考文献

[1] 白宇皓. 碳达峰, 碳中和背景下新能源发展趋势探讨[J].经济技术协作信息, 2021(35): 3. BAI Yuhao. Discussion on the development trend of new energy under the background of carbon peak and carbon neutrality[J]. Economic Technical Cooperation Infor-mation, 2021(35): 3.
[2] 曹阳, 孟茁. 基于无人机光伏巡检的视频图像传输系统设计[J]. 计算机与网络, 2020, 46(15):62-65. CAO Yang, MENG Zhuo. Design on video image transmission system based on uav photovoltaic inspection[J]. Computer & Network, 2020, 46(15): 62-65.
[3] 李香凡, 唐志伟, 胡家龙, 等. 无人机光伏巡检的图像目标检测算法[J]. 电子世界, 2020, 11: 7-8. LI Xaingfan, TANG Zhiwei, HU Jialong, et al. Image target detection algorithm for uav photovoltaic inspection[J]. Electronics World, 2020, 11: 7-8.
[4] SU B, ZHOU Z, CHEN H. PVEL-AD: a large-scale open-world dataset for photovoltaic cell anomaly detection[J]. IEEE Transactions on Industrial Informatics, 2022, 19(1): 404-413.
[5] 刘颖, 刘红燕, 范九伦, 等. 基于深度学习的小目标检测研究与应用综述[J]. 电子学报, 2020, 48(3): 590-601. LIU Ying, LIU Hongyan, FAN Jiulun, et al. A survey of research and application of small object detection based on deep learning[J]. Acta Electronica Sinica, 2020, 48(3): 590-601.
[6] 刘洋, 战荫伟. 基于深度学习的小目标检测算法综述[J]. 计算机工程与应用, 2021, 57(2):37-48. LIU Yang, ZHAN Yinwei. Survey of small object detection algorithms based on deep learning[J]. Journal of Computer Engineering & Applications, 2021, 57(2):37-48.
[7] WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C] //Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. Vancouver, Canada: IEEE, 2023: 7464-7475.
[8] FADA J S, HOSSAIN M A, BRAID J L, et al. Electroluminescent image processing and cell degradation type classification via computer vision and statistical learning methodologies[C] // Proceedings of IEEE 44th Photovoltaic Specialist Conference(PVSC). Washington, D.C., USA: IEEE, 2017: 3456-3461.
[9] KARIMI A M, FADA J S, HOSSAIN M A, et al. Automated pipeline for photovoltaic module electroluminescence image processing and degradation feature classification[J]. IEEE Journal of Photovoltaics, 2019, 9(5): 1324-1335.
[10] DEITSCH S, CHRISTLEIN V, BERGER S, et al. Automatic classification of defective photovoltaic module cells in electroluminescence images[J]. Solar Energy, 2019, 185: 455-468.
[11] SIMONYAN K, ZISSERMAN A. Very deep convol-utional networks for large-scale image recognition [EB/OL].(2015-04-10)[2024-07-15]. https://arxiv.org/abs/1409.1556
[12] BARTLER A, MAUCH L, YANG B, et al. Automated detection of solar cell defects with deep learning[C] //2018 26th European signal processing conference(EUSIPCO). Rome, Italy: IEEE, 2018: 2035-2039.
[13] TANG W, YANG Q, XIONG K, et al. Deep learning based automatic defect identification of photovoltaic module using electroluminescence images[J]. Solar Energy, 2020, 201: 453-460.
[14] AKRAM M W, LI G, JIN Y, et al. CNN based automatic detection of photovoltaic cell defects in electroluminescence images[J/OL]. Energy, 2019, 189: 116319 [2024-05-20]. https://doi.org/10.1016/j.energy.2019.116319
[15] CHEN X, KARIN T, JAIN A. Automated defect identification in electroluminescence images of solar modules[J]. Solar Energy, 2022, 242: 20-29.
[16] FIORESI J, COLVIN D J, FROTA R, et al. Automated defect detection and localization in photovoltaic cells using semantic segmentation of electroluminescence images[J]. IEEE Journal of Photovoltaics, 2021, 12(1): 53-61.
[17] SU B, CHEN H, CHEN P, et al. Deep learning-based solar-cell manufacturing defect detection with complementary attention network[J]. IEEE Transactions on Industrial informatics, 2020, 17(6): 4084-4095.
[18] MEETHAL A, GRANGER E, PEDERSOLI M. Cascaded zoom-in detector for high resolution aerial images[C] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver, Canada: IEEE, 2023: 2045-2054.
[19] LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C] //Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Hawaii, USA: 2017: 2117-2125.
[20] HONG M, LI S, YANG Y, et al. SSPNet: scale selection pyramid network for tiny person detection from uav images[J]. IEEE Geoscience and Remote Sensing Letters, 2021, 19: 1-5.
[21] HE J Y, CHENG Z Q, LI C, et al. Damo-streamnet: optimizing streaming perception in autonomous driving[EB/OL].(2023-05-20)[2024-07-15].https://arxiv.org/abs/2303.17144
[22] DU B, HUANG Y, CHEN J, et al. Adaptive sparse convolutional networks with global context enhancement for faster object detection on drone images[C] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver, Canada: IEEE, 2023: 13435-13444.
[23] SUN Y, WANG S, CHEN C, et al. Boundary-guided camouflaged object detection[C] // Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence. Vienna, Austria: IJCAI, 2022: 1335-1341.
[24] NOH J, BAE W, LEE W, et al. Better to follow, follow to be better: towards precise supervision of feature su-per-resolution for small object detection[C] //Proceedings of the IEEE/CVF International Conference on Computer Vision. California, USA: IEEE, 2019: 9725-9734.
[25] LIANG J, ZENG H, ZHANG L. Details or artifacts: a locally discriminative learning approach to realistic image super-resolution[C] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans, USA: IEEE, 2022: 5657-5666.
[26] XU C, WANG J, YANG W, et al. RFLA: gaussian receptive field based label assignment for tiny object detection[C] //European conference on computer vision. Cham, Switzerland: Springer, 2022: 526-543.
[27] LEI T, SUN R, WANG X, et al. CiT-Net: convolutional neural networks hand in hand with vision transformers for medical image segmentation[C] //Proceedings of the 32th International Joint Conference on Artificial Intelligence. Macao, China: IJCAI, 2023: 1017-1025.
[28] YU F, KOLTUN V. Multi-scale context aggregation by dilated convolutions[EB/OL].(2016-04-30)[2024-07-15]. https://arxiv.org/abs/1511.07122
[29] OUYANG D, HE S, ZHANG G, et al. Efficient multi-scale attention module with cross-spatial learning[C] //ICASSP 2023 IEEE International Conference on Acoustics, Speech and Signal Processing(ICASSP). Rhodes, Greek: IEEE, 2023: 1-5.
[30] ZHU X, LYU S, WANG X, et al. TPH-YOLOv5: improved yolov5 based on transformer prediction head for object detection on drone-captured scenarios[C] //Proceedings of the IEEE/CVF international conference on computer vision. Montreal, Canada: IEEE, 2021: 2778-2788.
[31] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C] //Proceedings of the IEEE International Conference on Computer Vision. Venice, Italy: IEEE, 2017: 2980-2988.
[32] ZHENG Z, WANG P, LIU W, et al. Distance-iou loss: faster and better learning for bounding box regression[C] //Proceedings of the AAAI conference on artificial intelligence. New York, USA: AAAI Press, 2020, 34(7): 12993-13000.
[33] NGUYEN D L, VO X T, PRIADANA A, et al. Car detector based on yolov5 for parking management[C] //Conference on Information Technology and its Applications. Cham, Switzerland: Springer, 2023: 102-113.
[34] 周颖,颜毓泽,陈海永,等. 基于改进YOLOv8的光伏电池缺陷检测[J]. 激光与光电子学进展, 2024, 61(8): 1-17. ZHOU Ying, YAN Yuze, CHEN Haiyong, et al. Defect detection of photovoltaic cells based on improved yolov8[J]. Laser & Optoelectronics Progress, 2024, 61(8): 1-17.
[35] ZHANG X, IZQUIERDO E, CHANDRAMOULI K. Dense and small object detection in uav vision based on cascade network[C] //Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops. Seoul, IEEE, 2019: 118-126.
[36] LI X, WANG W, WU L, et al. Generalized focal loss: learning qualified and distributed bounding boxes for dense object detection[J]. Advances in Neural Information Processing Systems, 2020, 33: 21002-21012.
[37] YANG C, HUANG Z, WANG N. Querydet: cascaded sparse query for accelerating high-resolution small object detection[C] //Proceedings of the IEEE/CVF Con-ference on Computer Vision and Pattern Recognition. New Orleans, USA: IEEE, 2022: 13668-13677.

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed