Artificial Intelligence for Pigment Classification Task in the Short-Wave Infrared Range.

Emeline Pouyet,Neda Rohani,Laurence De Viguerie,Tsveta Miteva

doi:10.3390/s21186150

Abstract

Hyperspectral reflectance imaging in the short-wave infrared range (SWIR, “extended NIR”, ca. 1000 to 2500 nm) has proven to provide enhanced characterization of paint materials. However, the interpretation of the results remains challenging due to the intrinsic complexity of the SWIR spectra, presenting both broad and narrow absorption features with possible overlaps. To cope with the high dimensionality and spectral complexity of such datasets acquired in the SWIR domain, one data treatment approach is tested, inspired by innovative development in the cultural heritage field: the use of a pigment spectral database (extracted from model and historical samples) combined with a deep neural network (DNN). This approach allows for multi-label pigment classification within each pixel of the data cube. Conventional Spectral Angle Mapping and DNN results obtained on both pigment reference samples and a Buddhist painting (thangka) are discussed.

Highlights

Works of art are created using a wide variety of materials, or mixtures of materials, and often exhibit heterogeneities at multiple length scales
The efficiency of the Spectral Angle Mapping Algorithm (SAM) and deep neural network (DNN) approaches to classify the different pixels of the The efficiency of the SAM and DNN approaches to classify the different pixels of the mockup sample dataset are compared in Figure 1c and 1d, respectively
The DNN approach proposed in this study offers new possibilities in the Short-Wave Infrared Hyperspectral Imaging (SWIR) range to identify and map pigments in complex materials either for unknown mixtures or multilayered systems

Summary

Introduction

Works of art are created using a wide variety of materials, or mixtures of materials, and often exhibit heterogeneities at multiple length scales. Their study requires the use of complementary analytical imaging techniques to probe and image the intrinsic complexity of such objects. The last few years saw a tremendous rise in hyperspectral reflectance imaging spectroscopy in the cultural heritage domain [3,4]. This success may be attributed to the merits of the technique: it is non-invasive, portable, and allows for the wide field imaging of an artwork in under a few minutes. Hyperspectral imaging in the visible range (400–900 nm) is a well-established technique to map the distribution of colorants across a painted surface.

Methods

Results

Conclusion