Centre for Sensors, Instruments and Systems Development

Universitat Politècnica de Catalunya

Shaping light to your needs

Foto Foto Foto Foto
   

News

12/01/2016
back

Gonio-hyperspectral imaging system based on light-emitting diodes for the analysis of automotive coatings

"Gonio-hyperspectral imaging system based on light-emitting diodes for the analysis of automotive coatings" is the PhD thesis that will present Francisco Javier Burgos on December 19th at the Joan Salvado Auditorium (Terrassa Optics and Optometry Faculty). The thesis has been directed by Meritxell Vilaseca and Jaume Pujol, researchers of the CD6 and faces the development of new colorimetric systems useful to measure coatings with their appearance depending on the illumination and observation angles. This type of applications are of special relevance in the automobile sector.


Venue: Joan Salvadó Auditorium (Terrassa Optics and Optometry Faculty)
Date: December 19th, 2016
Time: 10:00


Abstract:

Goniochromatic pigments, such as automotive coatings, are known by their appearance changes - especially of colour and visual texture - as a function of the angles of illumination and observation; they are classified as pearlescent (chroma and hue shifts) and metallic pigments (lightness variations). For this reason, their characterization through precise spectral and spatial data is of great importance in the industry. Accordingly, the main goal of this thesis is the development of a gonio-hyperspectral imaging system capable of acquiring spectral images with high spatial resolution using different angular configurations of illumination and observation. Conventional commercial systems used for this purpose provide accurate spectral information but they lack from spatial resolution as they use generally a photodiode and not an imaging sensor. The system has been designed to cover part of the ultraviolet (UV), the visible (VIS), and part of the infrared (IR) range of the electromagnetic spectrum. This thesis was primarily focused on the design, characterisation, validation and application of the system, which is based on light-emitting diodes (LEDs). lt includes a LED-based light source with 28 spectral bands from 368 nm to 1309 nm and one with white illumination, which can be automatically positioned in front of the sample; additionally, it has a CCD camera with sensitivity from 200 nm to 1000 nm and an lnGaAs camera with sensitivity from 900 nm to 1750 nm; a motorised rotation stage that controls the angle of observation by rotating the cameras around the sample and another that controls the angle of illumination by moving the sample are also used. 30 pearlescent, 30 metallic and 30 solid samples of commercial automotive coatings were characterized by means of the system to analyse their spectral reflectance and colour at different geometries, following the DIN 6175-2, ASTM E2194, and ASTM E2539 standards; the textural effects of sparkle, graininess and mottling were also evaluated. Purpose-built graphical user interfaces to control the hardware and processing the information obtained with the system were implemented. The spectral and colorimetric performance of the system was evaluated in the VIS using the colour-difference formulas CIEDE2000 and AUDl2000, the mean absolute error (MAE), the root mean square error (RMSE), and the goodness-of-fit coefficient (GFC) with respect to two commercial gonio spectrophotometers, the BYK-mac® and the X-Rite MA98®; the IR range was validated through the MAE, RMSE, and GFC with respect to the spectrometer INSTRUMENT SYSTEMS SPECTRO 320®. Regarding textural measurements, new specific indices - mainly based on first and second order statistics - have been developed in this thesis for the assessment of sparkle, graininess and mottling; the results of sparkle and graininess were compared with those of the BYK-mac® while the mottling assessment was only validated through visual observations. The system provided accurate results concerning spectral and colorimetric data for the VIS and IR range when compared with those of the reference devices. The largest differences were found for those geometries close to the specular reflection. Sparkle and graininess assessments also led to precise results, especially when applying the developed indices. Mottling results allowed to clearly differentiate between samples with and without this effect. The spectral analysis of these three textural descriptors also revealed very useful information that contributed with new measurement approaches. In conclusion, the results obtained in this thesis demonstrate the potential of the developed gonio-hyperspectral imaging system based on LEDs for the assessment of spectral, colorimetric and textural information of goniochromatic pigments in the UV, VIS and IR regions.
Extranet
CD6 Centre for Sensors, Instruments and Systems Development
Rambla de Sant Nebridi, 10  ·  08222  ·  Terrassa (Barcelona)