A new algorithm is presented for estimating the spectral reflectance of a
surface from a color image using highlights. The technique is based on
detailed reflectance models and is more general than existing techniques.
An estimate of spectral reflectance is useful for surface recognition
and characterization because it is intrinsic to the material, i.e. it
does not depend on illumination conditions. Experimental results obtained
by applying the algorithm to color images are presented.