LINEAR-MODELS FOR SPECTRAL REFLECTANCE FUNCTIONS OVER THE MID-WAVE AND LONG-WAVE INFRARED

We analyze the use of linear models for IR spectral reflectance functi ons. Linear models have been studied extensively for the visible wavel engths and form the basis of several approaches to estimating surface properties from color images. The IR analysis is performed with measur ed spectral reflectance functions for 394 samples of natural and man-m ade materials. The mid-wave (3-5 mu m) and long-wave (8-12.5 mu m) atm ospheric windows of the IR spectrum are considered separately. Since m aterials tend to have stronger spectral features over the 8-12.5 mu m range, linear models for the long-wave LR require more parameters than for the mid-wave IR, in order to obtain the same accuracy. We show th at a six-parameter linear model provides an excellent approximation fo r mid-wave (3-5 mu m) reflectance functions and that a nine-parameter linear model provides a satisfactory approximation for long-wave (8-12 .5 mu m) reflectance functions. (C) 1998 Optical Society of America.