Goodness of the reflectance prediction in multi-angle measurement of metallic and pearlescent samples mainly depends on the viewing and illumination directions. This study focuses on predicting the reflectance for all viewing directions from the reflectance for few number of selected best primary angles. The experiment was done on metallic and pearlescent samples. The principle component analysis and wiener estimation method were used to find best primary angles and predict the reflectance for all viewing angles. It has been found that wiener estimation method with higher polynomial order of Principle Components (PC) reduces prediction error significantly. Reflectance for only three best primary angles are sufficient to predict reflectance for all viewing angles between 125° to −35° in aspecular directions, if the estimation function was calculated from the training set of similar types. Similarly this experiment shows that reflectance for only five best primary angles are sufficient to predict reflectance for all viewing angles independent of the sample types.
Pesal Koirala, Markku Hauta-Kasari, Jouni Hiltunen, Jussi Parkkinen, "Reflectance Prediction in Multi-Angle Measurement by Wiener Estimation Method" in Proc. IS&T 16th Color and Imaging Conf., 2008, pp 221 - 226, https://doi.org/10.2352/CIC.2008.16.1.art00042