Large-Area Film Thickness Identification of Transparent Glass by Hyperspectral Imaging

• DOI: 10.3390/s24165094
• Publication Date: 2024-08-06T19:24:16Z
• Authors: Shuan-Yu Huang; Riya Karmakar; Yu-Yang Chen; Wei-Chin Hung; Arvind Mukundan; Hsiang-Chen Wang
• Keywords: Spectral bands; Chemical Imaging

This study introduces a novel method for detecting and measuring transparent glass sheets using hyperspectral imaging (HSI). The main goal of this is to create conversion technique that can accurately display spectral information from collected images, particularly in the visible light spectrum (VIS) near-infrared (NIR) areas. enables capture relevant data when used with images provided by industrial cameras. next step investigation principal component analysis examine obtained derived different treated samples. analytical procedure standardizes magnitude wavelengths are inherent HSI images. simulated profiles generalized inverse matrix on normalized These then matched spectroscopic microscopic imaging, resulting observation distinct dispersion patterns. use coloring methods effectively displays thickness processing sheet visually noticeable way. Based empirical research, changes coating NIR-HSI range cause significant transmission infrared at within NIR spectrum. phenomenon serves as foundation film thickness. root mean square error inside area impressively low, calculated be just 0.02. highlights high level accuracy achieved stated above. Potential areas arise incorporating proposed approach into design real-time, wide-scale automated optical inspection system.