Image-Based Bidirectional Reflectance Measurement of Non-Diffuse and Gonio-Chromatic Materials

Abstract

Visual appearance of packaging material can vary in terms of perceived colour or lightness properties with a change in illumination and viewing directions mainly due to its optical properties. This gives a desirable visual appearance and makes the package look attractive and standout on a supermarket or an airport duty-free shelf. To characterise and reproduce such a material, objective measurements in terms of the amount of the light incident and reflected from the material surface are performed. To perform such measurements, standardisation bodies like the International Commission on Illumination (CIE) and the ASTM International recommended measurement geometries based on the material optical properties. Single measurement geometries recommended by these standardisation bodies, however, are inadequate to measure and characterise materials that reflect the incident light in a non-diffuse and goniometric way. Bidirectional measurements are required to characterise such packaging print materials which can be difficult and time consuming. Gonio-spectrophotometers that measure at a broad number of illumination and viewing directions are commercially available and used to perform bidirectional reflectance measurements. Gonio-spectrophotometers can be slow due to moving light source, to illuminate a flat measurement sample, and a detector, to record the light reflected from the sample surface. To overcome these drawbacks, image-based measurement techniques have been proposed and presented in the past. An image-based measurement setup uses a camera as a detector to objectively measure the incident and reflected light from the material surface. In this thesis we investigate the applicability of using an image-based measurement technique to perform bidirectional reflectance measurements, analytically estimate, and represent the bidirectional reflectance distribution function (BRDF) of flexible and homogeneous packaging print materials having different optical properties. We present an image-based measurement setup (measurement setup) to perform fast bidirectional reflectance measurements of flexible, homogeneous packaging print materials. The packaging materials measured in this thesis show reflectance properties that vary from diffuse to non-diffuse and gonio-chromatic. We evaluate the accuracy of the measurement setup by comparing it against two commercially available gonio-spectrophotometers. The uncertainty in calculating the incident and viewing directions was large and is dependent on the physical measurements within the measurement setup. The practical use of such a setup to perform bidirectional reflectance measurements is dependent on the material to be measured, precision of the physical measurements within the measurement setup and their implications. Further we investigate the suitability of using the measurement setup to perform bidirectional reflectance measurements and analytically estimate material BRDF using an analytical reflectance model. Measurements from two commercially available gonio-spectrophotometers are used to validate the measurement setup. Depending on the material reflectance properties, different combinations of reflectance models, measurement datasets and cost functions are used for estimation. A salient measurement dataset is investigated for BRDF estimation and representation of the materials measured in this thesis. Retro-reflective measurements obtained using the measurement setup provided a good BRDF estimation for visual representation. Measurements in addition to retro-reflective measurements did improve the visual representation of the material with gonio-chromatic relfectance property. The measurement setup can be used as a tool to generate BRDF datasets of flexible, homogeneous, and isotropic materials having different reflectance properties provided precise physical measurements within the measurement setup are performed.