Solar cell defect detection based on improved G-SSD Network
DOI:
https://doi.org/10.54097/ije.v2i1.5618Keywords:
Solar cell defect detection, G-SSD NetworkAbstract
Aiming at the problems of finger interruptions, microcracks and shadow defects of solar cells that are not easy to detect and will reduce the efficiency of the cells, a solar cell defect localization detection algorithm based on improved G-SSD network is proposed. For object detection algorithms, the detection of small targets is challenging. Firstly, the input scale and network model are improved for the detection of small-scale targets. Secondly, in order to reduce the weight of the network model, a more lightweight network GhostNet is selected instead of VGG-16 in the experiment to reduce the computational cost of the network model. Finally, the EL image is input into the network for classification and positioning, and the prediction results are integrated and the final detection results are output. Experimental results show that compared with the original model, the mean (mAP) of the improved G-SSD is increased by 0.68%, effectively reduces some computing costs.
Downloads
References
S. Ostapenko, W. Dallas, D Hess, et al. Crack Detection and Analyses using Resonance Ultrasonic Vibrations in Crystalline Silicon Wafers; proceedings of the IEEE World Conference on Photovoltaic Energy Conversion, F, 2007 [C].
D. Van et al. Optimizing the front side metallization procems using the corescan [J]. Conference record of the twenty-ninth IEEE photovoltaic specialists conference,2002, 340-3.
T. Trupke, R.A. Bardos, M. Abbott, et al. Progress with luminescence imaging for the characterisation of silicon wafers and solar cells [J]. 2007.
Karsten Bothe P, Jan Schmidt. Electrolumines ceflce imaging as an in-line characterizati On tool for solar cell production [J]. Record of the 21st European Photovoltaic Solar, 2006, 597-660.
AY Liu. Research on weak boundary defect detection technology of solar cells based on deep learning [D]; Wuhan University of Science and Technology, 2021.
XJ Tian. Research on silicon wafer crack detection algorithm based on optimized YOLOv3 neural network [D]; North University of China, 2019.
DX Wang. Research on multi-target detection and tracking method based on deep learning. Dalian University of Technology [J]. 2021,
Misra D. Mish: A Self Regularized Non-Monotonic Neural Activation Function [M]. 2019.
Buerhop-Lutz C, Winkler T, Fecher F W, et al. Performance Analysis of Pre-Cracked PV-Modules at Realistic Loading Conditions; proceedings of the 33rd EUPVSEC, F, 2017 [C].
Downloads
Published
Issue
Section
How to Cite
