Spectral reflectance characterization and fiber type discrimination for common natural textile materials using a portable spectroradiometer

Publication date: November 2019

Source: Journal of Archaeological Science, Volume 111

Author(s): Hengqian Zhao, Yunli Wang, Shuai Liu, Kunheng Li, Wei Gao

Abstract

Component identification of textile materials is an important issue for studies of cultural heritage collections and science conservation. Portable spectroradiometer technology as a potential technique can be utilized to discriminate mass natural fibers rapidly and nondestructively. The goal of this study was to examine the feasibility of portable spectroradiometer to natural fiber type discrimination in the spectral range covering visible and near infrared (VNIR) (350–1000 nm), shortwave infrared 1 (SWIR1) (1000–1850 nm) and SWIR2 (1850–2500 nm). For this purpose, reflection spectra of four types of natural fibers commonly used in archaeological textiles were measured using a portable spectroradiometer under standardized conditions. The spectral features of four categories of natural fibers in different wavelength ranges were extracted and compared. To further explore the potential of discriminating the fiber types using portable spectroradiometer, principal component analysis in combination with statistical tests was applied. The results indicated that fiber type had a strong influence on the spectral features, which was much more remarkable than the influences of the intra-class variation and physical properties. Moreover, with large inter-class variation and small intra-class variation, SWIR2 was more suitable for fiber type identification than VNIR and SWIR1, which was also validated in the statistical analysis. Besides, the approach of selecting effective principal components in different wavelength ranges could provide a more optimized combination of differential information for the correct assignment of the investigated natural fiber types of textile materials. In general, portable spectroradiometer could be a powerful technique for on-site fiber type discrimination.