Extended hyperspectral characterization of plastic automotive parts via Acousto-optic Tunable Filter and Fourier Transform Infrared Spectrometry

Abstract

The importance of plastics in the design and construction of a new generation of vehicles is fundamental for the automotive industry. Typically characterization of these components is done by using techniques such as Fourier Transform Infrared Spectrometer (FTIR), and spectrometer in the visible bands (SPM), among others. It is also possible to get spectral information with high resolution using Acousto-optic Tunable Filters (AOTF). Since the characterization of plastic or polymer materials has been studied extensively, one problem is still open in order to be understood - Cosmetic Degradation. Particularly in the automotive industry, this information is fundamental to guarantee the quality and security of components that are used to build vehicles. Considering the above, we propose a new strategy based on an extended hyperspectral characterization of plastics by using transform and juxtaposition operations of reflectance information values gathered from three techniques: FTIR, AOTF, and SPM. Experimental results show that the proposed method contributes significantly to understanding the problem of cosmetic degradation in plastics, and opens a new branch to understanding the phenomenon through modern computational techniques based on artificial intelligence.