针对太阳电池片EL图像,提出一种融合注意力机制和Ghost卷积层并引入批标准化的T-VGG轻量级卷积神经网络模型。首先使用Ghost卷积层替换常规卷积层,其次引入注意力机制和批规范化,进而实现对电池片缺陷的高精高速检测。实验结果表明,该卷积神经网络模型对缺陷的检测准确率为99.15%,对缺陷类型的检测准确率为96.28%,检测时间为0.032 s/张,在保证高精高效性的同时兼具通用性。
Abstract
A light-weight convolutional neural network model with batch standardized T-VGG （Tiny Visual Geometry Group） was proposed to integrate attention mechanism and Ghost block into the EL image of solar cells. Using of Ghost convolutional layer to replace the conventional convolutional layer, followed by the introduction of attention and batch standardization, so as to achieve high precision and high-speed detection of battery defects. The experimental results show that the accuracy of the convolutional neural network model for defect detection is 99.15%, The detection accuracy of defect type is 96.28%, and the time is 0.032 s/ piece, which not only ensures high precision and high efficiency, but also has universality.
关键词
深度学习 /
卷积神经网络 /
太阳电池 /
缺陷检测 /
T-VGG
Key words
deep learning /
convolutional neural network /
solar cells /
defect detection /
T-VGG
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 范程华, 王群京, 曹欣远, 等. 基于欠定方程的太阳电池片表面缺陷检测算法[J]. 太阳能学报, 2020, 41(6): 288-292.
FAN C H, WANG Q J, CAO X Y, et al.Surface defect detection algorithm for Solar cell Based on underdetermined equation[J]. Acta energiae solaris sinica, 2020, 41(6): 288-292.
[2] ANWAR S A, ABDULLAH M Z.Micro-crack detection of multicrystalline solar cells featuring an improved anisotropic diffusion filter and image segmentation technique[J]. Eurasip journal on image & video processing, 2014, 2014(1): 1-17.
[3] 龚芳, 张学武, 孙浩. 基于独立分量分析和粒子群算法的太阳能电池表面缺陷红外热成像检测[J]. 光学学报, 2012, 32(4): 169-177.
GONG F, ZHANG X W, SUN H.Infrared thermal imaging detection of surface defects in solar cells based on independent component analysis and particle swarm optimization[J]. Journal of optics, 2012, 32(4): 169-177.
[4] TSAI D M,WU S C,CHIU W Y.Defect detection in solar modules using ICA basis images[J]. IEEE transactions on industrial informatics, 2013, 9(1): 122-131.
[5] XU P, ZHOU W J, FEI M R.Detection methods for micro-cracked defects of photovoltaic modules based on machine vision[C]//The 2014 IEEE 3rd International Conference on Cloud Computing and Intelligence Systems(CCIS), New York, USA, IEEE, 2014: 609-613.
[6] 墨恺, 徐林. 基于阈值均匀局部二值模式和BP神经网络的太阳电池缺陷检测算法[J]. 太阳能学报, 2014, 35(12): 2448-2454.
MO K, XU L.Defect detection algorithm for solar cells based on threshold uniform local binary model and bp neural network[J]. Acta energiae solaris sinica, 2014, 35(12): 2448-2454.
[7] 陶显, 侯伟, 徐德. 基于深度学习的表面缺陷检测方法综述[J]. 自动化学报, 2021, 47(5): 1017-1034.
TAO X, HOU W, XU D.A Review of surface defect detection methods based on deep learning[J]. Acta automatica sinca, 2021, 47(5): 1017-1034.
[8] HUANG G, LIU Z, VAN DER M L, et al. Densely connected convolutional networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR), Honolulu, HI, USA, 2017.
[9] LIU W, ANGUELOV D, ERHAN D, et al.SSD: Single shot multibox detector[C]//Proceedings of European Conference on Computer Vision(ECCV), Amsterdam, The Netherlands, Springer, 2016.
[10] REDMON J, FARHADI A.YOLOV3: An incremental improvement[C]//Proceedings of 2018 IEEE Conference on Computer Vision and Pattern Recognition(CVPR), Washington DC, USA, IEEE Press, 2018.
[11] 刘怀广, 刘安逸, 周诗洋, 等. 基于深度神经网络的太阳能电池组件缺陷检测算法研究[J]. 应用光学, 2020, 41(2): 327-336.
LIU H G, LIU X Y, ZHOU S Y, et al.Research on defect detection algorithm of solar cell module based on deep neural network[J]. Applied optics, 2020, 41(2): 327-336.
[12] HU J, SHEN L, SUN G, et al.Squeeze-and-excitation networks[C]//Proceedings of the 30th IEEE Recognition Conference on Computer Vision and Pattern (CVPR), Piscataway, NJ, USA, IEEE, 2017.
[13] HAN K, WANG Y H, TIAN Q, et al.More features from cheap operations[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 2020.
[14] TAN M X, LE Q V.EfficientNet: rethinking model scaling for convolutional neural net-works[C]//International Conference on Machine Learning (ICML), Long Beach, CA, USA, 2019: 6105-6114.
[15] SZEGEDY C, VANHOUCKZ V, IOFFE S, et al.Rethinking the Inception Architecture for Computer Vision[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 2016.
基金
国家重点研发计划（2018YFB1403303）
PDF(2649 KB)
