针对以往光伏缺陷检测中可识别缺陷的种类少、无法对缺陷进行定位、模型参数多体积大以及检测速度慢的局限性,改进传统的YOLOv5网络对光伏组件面板图像中常见的裂纹、断栅、黑芯、粗线和热斑5类主要缺陷进行检测和分类。使用3种不同的注意力机制模块：CA注意力机制模块、ECA注意力机制模块、CBAM注意力机制模块,分别融入YOLOv5网络中进行对比分析实验,发现CA注意力机制更加适合光伏缺陷图像检测。随后对融入CA注意力机制模块的YOLOv5算法再次加入双向特征金字塔网络结构,进一步加强网络的特征融合能力。实验结果表明,该模型可对5类常见的缺陷进行有效的识别和定位,与YOLOv5算法相比平均精准度（mAP）值提升3.7 %,模型体积减小15 %,图片的检测平均速度提升9.7 %。结论表明该方法可有效增强YOLOv5算法在光伏缺陷检测中的能力,同时可降低深度学习算法在光伏检测中误检和漏检的情况。
Abstract
Aiming at the limitations of the previous PV defect detection, such as fewer types of recognizable defects, multiple model parameters, large volume of model parameters, and slow detection speed, the traditional YOLOv5 network is improved to detect and classify the five main types of defects, namely, cracks, broken grids, black cores, thick wires, and hot spots that are commonly found in the images of photovoltaic panels. Three different attention mechanism modules： CA attention mechanism module, ECA attention mechanism module, and CBAM attention mechanism module, are integrated into the YOLOv5 network for comparative analysis experiments, and it is found that the CA attention mechanism is more suitable for PV defect image detection. Subsequently, the YOLOv5 algorithm incorporating the CA attention mechanism module is added to the bidirectional feature pyramid network structure further to strengthen the feature fusion capability of the network. The experimental results show that the model can effectively identify and localize five types of common defects, and compared with the YOLOv5 algorithm, the Map （Mean Average Precision） value is improved by 3.7 %, the model volume is reduced by 15 %, and the average speed of the detection of the images is improved by 9.7 %. The overall conclusion shows that the method effectively enhances the ability of YOLOv5 algorithm in PV defect detection, and at the same time reduces the misdetection and omission of the deep learning algorithm in PV detection.
关键词
计算机视觉 /
深度学习 /
太阳电池 /
YOLOv5 /
光伏缺陷
Key words
computer vision /
deep learning /
solar cells /
YOLOv5 /
photovoltaic defects
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 丁世浩. 基于计算机视觉的光伏组件缺陷诊断研究[D].杭州:浙江大学, 2020.
DING S H.Research on defect diagnosis of photovoltaic module based on computer vision[D]. Hangzhou: Zhejiang University, 2020.
[2] 蔡洁聪, 吕洪坤, 朱凌云, 等. 光伏发电站热斑检测技术综述[J]. 电源技术, 2021, 45(5): 683-685.
CAI J C, LYU H K, ZHU L Y, et al.Review of hot spot detection technology in photovoltaic power station[J]. Chinese journal of power sources, 2021, 45(5): 683-685.
[3] 郭宝柱. 光伏阵列热斑的红外图像处理的研究[D].天津: 天津理工大学, 2016.
GUO B Z.Study on infrared image processing of hot spots in photovoltaic array[D]. Tianjin: Tianjin University of Technology, 2016.
[4] 蒋琳, 苏建徽, 施永, 等. 基于红外热图像处理的光伏阵列热斑检测方法[J]. 太阳能学报, 2020, 41(8): 180-184.
JIANG L, SU J H, SHI Y, et al.Hot spots detection of operating PV arrays through IR thermal image[J]. Acta energiae solaris sinica, 2020, 41(8): 180-184.
[5] 马浩. 基于红外热成像及可见光图像的光伏电站热斑检测及定位技术研究[D]. 南京: 南京邮电大学, 2019.
MA H.Research on hot spot detection and localization technology of photovoltaic power station based on infrared thermal imaging and visible light image[D]. Nanjing: Nanjing University of Posts and Telecommunications, 2019.
[6] HE K M, GKIOXARI G, DOLLÁR P, et al. Mask R-CNN[C]//2017 IEEE International Conference on Computer Vision (ICCV). Venice, Italy, 2017: 2980-2988.
[7] 王道累, 李超, 李明山, 等. 基于深度卷积神经网络的光伏组件热斑检测[J]. 太阳能学报, 2022, 43(1): 412-417.
WANG D L, LI C, LI M S, et al.Solar photovoltaic modules hot spot detection based on deep convolutional neural networks[J]. Acta energiae solaris sinica, 2022, 43(1): 412-417.
[8] LIU W, ANGUELOV D, ERHAN D, et al.SSD: single shot MultiBox detector[M]. Lecture Notes in Computer Science, Cham: Springer International Publishing, 2016: 21-37.
[9] REDMON J, DIVVALA S, GIRSHICK R, et al.You only look once: Unified, real-time object detection[C]//Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas, NV, USA, 2016: 779-788.
[10] 孙海蓉, 李帆. 基于注意力机制的光伏热斑识别[J]. 太阳能学报, 2023, 44(2): 453-459.
SUN H R, LI F.Photovoltaic hot spot recognition based on attention mechanism[J]. Acta energiae solaris sinica, 2023, 44(2): 453-459.
[11] 郭岚, 刘正新. 基于改进YOLOv5的光伏组件缺陷检测[J]. 激光与光电子学进展, 2023, 60(20): 148-156.
GUO L, LIU Z X.Improved YOLOv5-based defect detection in photovoltaic modules[J]. Laser & optoelectronics progress, 2023, 60(20): 148-156.
[12] YAN B, FAN P, LEI X Y, et al.A Real-Time apple targets detection method for picking robot based on improved YOLOv5[J]. Remote Sensing, 2021, 13(9) : 1619.
[13] DAI J, ZHAO X, LI L P, et al.Improved YOLOv5-based infrared dim-small target detection under complex background[J]. Infrared technology, 2022, 44(5): 504-512.
[14] SENG W C, MIRISAEE S H.A new method for fruits recognition system[C]//2009 International Conference on Electrical Engineering and Informatics. Bangi, Malaysia, 2009: 130-134.
[15] LINKER R, COHEN O, NAOR A.Determination of the number of green apples in RGB images recorded in orchards[J]. Computers and electronics in agriculture, 2012, 81(1): 45-57.
[16] KURTULMUS F, LEE W S, VARDAR A.Immature peach detection in colour images acquired in natural illumination conditions using statistical classifiers and neural network[J]. Precision agriculture, 2014, 15(1): 57-79.
[17] 张志远, 罗铭毅, 郭树欣, 等. 基于改进YOLOv5的自然环境下樱桃果实识别方法[J]. 农业机械学报, 2022, 53(S1): 232-240.
ZHANG Z Y, LUO M Y, GUO S X, et al.A cherry fruit recognition method based on improved YOLOv5 in natural environment[J]. Journal of agricultural machinery, 2022, 53(S1): 232-240.
[18] MOHAMMED S, NIMALI T.A new paradigm for waste classification based on YOLOv5[J]. Instrumentation, 2021, 8(4): 9-17.
[19] 彭继慎, 孙礼鑫, 王凯, 等. 基于模型压缩的ED-YOLO电力巡检无人机避障目标检测算法[J]. 仪器仪表学报, 2021, 42(10): 161-170.
PENG J S, SUN L X, WANG K, et al.ED-YOLO power inspection UAV obstacle avoidance target detectionalgorithm based on model compression[J]. Chinese journal of scientific instrument, 2021, 42(10): 161-170.
[20] GUO M H, XU T X, LIU J J, et al.Attention mechanisms in computer vision: a survey[J]. Computational visual media, 2022, 8(3): 331-368.
基金
中国高校产学研创新基金-新一代信息技术创新重点项目（2022IT017）; 中国劳动电子学会“产教融合,校企合作”教育改革发展课题（Ceal2024104）; 天津农学院人才资助计划（Y0400907）; 天津农学院实验教学和教学实验室建设研究项目（2024-3-01）
PDF(1333 KB)
