A threshold optimization strategy for reducing the noise of co-planar array capacitive imaging. (January 2023)

• Record Type: Journal Article
• Authors: Li, Ruihang; Zhang, Yuyan; Wen, Yintang; Pan, Zhao; Zhao, Zhenhao
• Is Part Of: NDT & E international. Volume 133(2023)
• Publication Date: 2023-01
• Subjects: Non-destructive testing -- Co-planar array capacitive imaging -- Threshold optimization -- Image reconstruction; Nondestructive testing -- Periodicals; Contrôle non destructif -- Périodiques; Electronic journals; 620.1127
• DOI: 10.1016/j.ndteint.2022.102738
• Languages: English
• ISSNs: 0963-8695

Abstract

Abstract: Co-planar array capacitive imaging is a non-destructive testing technique based on the sensing of dielectric properties. As the number of independent electrodes in the sensor grows, handling noise becomes more and more critical. In this paper, numerical models and practical tests were used to explore the noise analysis and processing of the output signals of a 12-electrode co-planar array capacitive sensor. The effects of different geometric placements and distances on the output signal strength and sensitivity of electrode pairs were thoroughly assessed, revealing that distal interpolar spacing electrode pairs were particularly vulnerable to noise. As a consequence, a threshold optimization strategy was presented for decreasing the noise. Results revealed that threshold optimization efficiently reduced noise and emphasized the useful information for image reconstruction. The artifacts and deviations of the reconstructed images were reduced, and the structural similarity and Pearson product-moment correlation coefficients were boosted. The methodologies and findings may provide a reference for signal processing in co-planar array capacitive imaging. Graphical abstract: Image 1 Highlights: The output signals of the far-spaced electrode pairs may contain noise. The gain signals were obscured by these sounds after normalization. The proposed threshold optimization approach effectively solved this issue. The bias and artifacts of the reconstructed images were effectively reduced. Threshold optimization is a necessary pre-processing step before image reconstruction.