A temporal and spatial EWMA-based monitoring and diagnosing approach for image data

ABSTRACT

With the increasing application of machine vision systems, it is crucial to monitor images of industrial products for potential anomalies in the production process. In current literature, the research interest mainly lies in either detection of fault occurrence or identification of fault size/location. Simultaneous fault detection and identification, however, have received limited attention. This paper focuses on fault detection and identification based on images of products whose quality is characterized by uniformity or conformity to a specific pattern. A temporal and spatial EWMA-based control charting method is proposed, where spatial correlation is incorporated to efficiently screen out the suspicious region. Results from extensive simulations show that the proposed control chart is superior especially in detecting and estimating small-area faults, and performs competitively well in monitoring and diagnosing large-area faults. An experimental case is analyzed to illustrate the application of the proposed method in practice.

KEYWORDS:

• Machine vision system
• image data
• statistical process control
• temporal and spatial smoothing
• EWMA-based statistic

Acknowledgements

The authors sincerely thank the Editor and the three anonymous reviewers for their thoughtful and constructive comments. This work was supported by the Doctoral Scientific Research Foundation of Shenzhen Institute of Information Technology under Grant [number SZIIT2021KJ004]; the National Natural Science Foundation of China under Grant [numbers 72072080, 72032005, 71672122, 71872123]; the General Project of Philosophy and Social Science Research in Colleges and Universities of Jiangsu Province under Grant [number 2021SJA0299]; and the Young Scholars Supporting Program of Nanjing University of Finance and Economics.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/16843703.2024.2321813

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [72072080, 72032005, 71672122, 71872123]; the Young Scholars Supporting Program of Nanjing University of Finance and Economics; General Project of Philosophy and Social Science Research in Colleges and Universities of Jiangsu Province [2021SJA0299]; Doctoral Scientific Research Foundation of Shenzhen Institute of Information Technology [SZIIT2021KJ004].

Notes on contributors

Ling Zuo

Ling Zuo is an assistant professor in the School of Management at Shenzhen Institute of Information Technology in China. She received her PhD in management science from Tianjin University in 2016. Her research interests include SPC and quality improvement. Her research appears in International Journal of Production Research, Journal of Intelligent Manufacturing, Quality Technology & Quantitative Management, and Journal of Systems Engineering (Chinese).

Panpan Zhou

Panpan Zhou is an assistant professor in the School of Management Science and Engineering at Nanjing University of Finance and Economics in China. She received her PhD in management science from Tianjin University in 2019 and visited Pennsylvania State University from December 2015 to December 2017. Her research interests include SPC and quality improvement. Her research appears in Computers & Industrial Engineering, Quality and Reliability Engineering International etc.