Study on performance evaluation of intelligent photovoltaic defect detection by electroluminescence imaging

Wende Liu; Yadong Chen; Taotao Zhang; Zuo Chen; Haiyong Gan; Junchao Zhang; Chuan Cai; Limin Xiong

doi:10.1117/12.3005643

Study on performance evaluation of intelligent photovoltaic defect detection by electroluminescence imaging

Wende Liu^*
, Yadong Chen
, Taotao Zhang^*
, Zuo Chen
, Haiyong Gan
, Junchao Zhang
, Chuan Cai
, Limin Xiong

^*Corresponding author for this work

School of Transportation Science and Engineering

National Institute of Metrology China
Beihang University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Machine vision is not a mere upgrade of the specification of the current imaging devices, but rather a form of visual perception technology that involves intelligent modules in the processes of measurement, processing, and decision- making. Given the novel functionalities and features of machine vision-based intelligent detection devices, the traditional evaluation methods based on testing the physical parameters of imaging devices need further refinement and development. Taking the electroluminescence (EL) imaging in photovoltaic (PV) tests as an example, we investigate the influence of changes in dataset characteristics on the performance of object detection by combining digital image processing and deep learning methods. Features regarding to the crack-type defect datasets, such as the grayscale, contrast, shape and resolution, are controlled and adjusted based on new generated datasets from the original datasets. From the numerical experiments, some new aspects for evaluating the intelligent detection.

Original language	English
Title of host publication	AOPC 2023
Subtitle of host publication	AI in Optics and Photonics
Editors	Juejun Hu, Jianji Dong
Publisher	SPIE
ISBN (Electronic)	9781510672383
DOIs	https://doi.org/10.1117/12.3005643
State	Published - 2023
Event	2023 Applied Optics and Photonics China: AI in Optics and Photonics, AOPC 2023 - Beijing, China Duration: 25 Jul 2023 → 27 Jul 2023

Publication series

Name	Proceedings of SPIE - The International Society for Optical Engineering
Volume	12966
ISSN (Print)	0277-786X
ISSN (Electronic)	1996-756X

Conference

Conference	2023 Applied Optics and Photonics China: AI in Optics and Photonics, AOPC 2023
Country/Territory	China
City	Beijing
Period	25/07/23 → 27/07/23

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

deep learning
defect detection
electroluminescence imaging
machine vision
performance evaluation
photovoltaic

Access to Document

10.1117/12.3005643

Cite this

Liu, W., Chen, Y., Zhang, T., Chen, Z., Gan, H., Zhang, J., Cai, C., & Xiong, L. (2023). Study on performance evaluation of intelligent photovoltaic defect detection by electroluminescence imaging. In J. Hu, & J. Dong (Eds.), AOPC 2023: AI in Optics and Photonics Article 129660L (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 12966). SPIE. https://doi.org/10.1117/12.3005643

@inproceedings{6da8130f4377428eb8a08035dc86dcd4,

title = "Study on performance evaluation of intelligent photovoltaic defect detection by electroluminescence imaging",

abstract = "Machine vision is not a mere upgrade of the specification of the current imaging devices, but rather a form of visual perception technology that involves intelligent modules in the processes of measurement, processing, and decision- making. Given the novel functionalities and features of machine vision-based intelligent detection devices, the traditional evaluation methods based on testing the physical parameters of imaging devices need further refinement and development. Taking the electroluminescence (EL) imaging in photovoltaic (PV) tests as an example, we investigate the influence of changes in dataset characteristics on the performance of object detection by combining digital image processing and deep learning methods. Features regarding to the crack-type defect datasets, such as the grayscale, contrast, shape and resolution, are controlled and adjusted based on new generated datasets from the original datasets. From the numerical experiments, some new aspects for evaluating the intelligent detection.",

keywords = "deep learning, defect detection, electroluminescence imaging, machine vision, performance evaluation, photovoltaic",

author = "Wende Liu and Yadong Chen and Taotao Zhang and Zuo Chen and Haiyong Gan and Junchao Zhang and Chuan Cai and Limin Xiong",

note = "Publisher Copyright: {\textcopyright} COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.; 2023 Applied Optics and Photonics China: AI in Optics and Photonics, AOPC 2023 ; Conference date: 25-07-2023 Through 27-07-2023",

year = "2023",

doi = "10.1117/12.3005643",

language = "英语",

series = "Proceedings of SPIE - The International Society for Optical Engineering",

publisher = "SPIE",

editor = "Juejun Hu and Jianji Dong",

booktitle = "AOPC 2023",

address = "美国",

}

Liu, W, Chen, Y, Zhang, T, Chen, Z, Gan, H, Zhang, J, Cai, C & Xiong, L 2023, Study on performance evaluation of intelligent photovoltaic defect detection by electroluminescence imaging. in J Hu & J Dong (eds), AOPC 2023: AI in Optics and Photonics., 129660L, Proceedings of SPIE - The International Society for Optical Engineering, vol. 12966, SPIE, 2023 Applied Optics and Photonics China: AI in Optics and Photonics, AOPC 2023, Beijing, China, 25/07/23. https://doi.org/10.1117/12.3005643

Study on performance evaluation of intelligent photovoltaic defect detection by electroluminescence imaging. / Liu, Wende; Chen, Yadong; Zhang, Taotao et al.
AOPC 2023: AI in Optics and Photonics. ed. / Juejun Hu; Jianji Dong. SPIE, 2023. 129660L (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 12966).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Study on performance evaluation of intelligent photovoltaic defect detection by electroluminescence imaging

AU - Liu, Wende

AU - Chen, Yadong

AU - Zhang, Taotao

AU - Chen, Zuo

AU - Gan, Haiyong

AU - Zhang, Junchao

AU - Cai, Chuan

AU - Xiong, Limin

PY - 2023

Y1 - 2023

N2 - Machine vision is not a mere upgrade of the specification of the current imaging devices, but rather a form of visual perception technology that involves intelligent modules in the processes of measurement, processing, and decision- making. Given the novel functionalities and features of machine vision-based intelligent detection devices, the traditional evaluation methods based on testing the physical parameters of imaging devices need further refinement and development. Taking the electroluminescence (EL) imaging in photovoltaic (PV) tests as an example, we investigate the influence of changes in dataset characteristics on the performance of object detection by combining digital image processing and deep learning methods. Features regarding to the crack-type defect datasets, such as the grayscale, contrast, shape and resolution, are controlled and adjusted based on new generated datasets from the original datasets. From the numerical experiments, some new aspects for evaluating the intelligent detection.

AB - Machine vision is not a mere upgrade of the specification of the current imaging devices, but rather a form of visual perception technology that involves intelligent modules in the processes of measurement, processing, and decision- making. Given the novel functionalities and features of machine vision-based intelligent detection devices, the traditional evaluation methods based on testing the physical parameters of imaging devices need further refinement and development. Taking the electroluminescence (EL) imaging in photovoltaic (PV) tests as an example, we investigate the influence of changes in dataset characteristics on the performance of object detection by combining digital image processing and deep learning methods. Features regarding to the crack-type defect datasets, such as the grayscale, contrast, shape and resolution, are controlled and adjusted based on new generated datasets from the original datasets. From the numerical experiments, some new aspects for evaluating the intelligent detection.

KW - deep learning

KW - defect detection

KW - electroluminescence imaging

KW - machine vision

KW - performance evaluation

KW - photovoltaic

UR - https://www.scopus.com/pages/publications/85181149121

U2 - 10.1117/12.3005643

DO - 10.1117/12.3005643

M3 - 会议稿件

AN - SCOPUS:85181149121

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - AOPC 2023

A2 - Hu, Juejun

A2 - Dong, Jianji

PB - SPIE

T2 - 2023 Applied Optics and Photonics China: AI in Optics and Photonics, AOPC 2023

Y2 - 25 July 2023 through 27 July 2023

ER -

Study on performance evaluation of intelligent photovoltaic defect detection by electroluminescence imaging

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this