Hyperspectral Imaging for Moisture Analysis

This work investigates the use of hyperspectral imaging (HSI) for the non-contact estimation of moisture content in materials of different natures, with applications in industrial drying processes. Hyperspectral data acquired in the SWIR range (1000-2500 nm) enabled the identification of water-sensitive wavelengths (mainly 1490 nm and 1980 nm) and the definition of spectral parameters, including reflectance ratios, correlated with gravimetrically measured moisture content. Principal Component Analysis (PCA) revealed a monotonic relationship between selected components and moisture content. Furthermore, a methodology for generating spatially resolved moisture maps was developed. The results demonstrate the feasibility of HSI as a reliable, non-contact, technology for moisture monitoring and quality control in industrial applications.