SEPARATION OF REFLECTION COMPONENTS USING COLOR AND POLARIZATION

Specular reflections and interreflections produce strong highlights in brightness images. These highlights can cause vision algorithms for s egmentation, shape from shading, binocular stereo, and motion estimati on to produce erroneous results. A technique is developed for separati ng the specular and diffuse components of reflection from images. The approach is to use color and polarization information, simultaneously, to obtain constraints on the reflection components at each image poin t. Polarization yields local and independent estimates of the color of specular reflection. The result is a linear subspace in color space i n which the local diffuse component must lie. This subspace constraint is applied to neighboring image points to determine the diffuse compo nent. In contrast to previous separation algorithms, the proposed meth od can handle highlights on surfaces with substantial texture, smoothl y varying diffuse reflectance, and varying material properties. The se paration algorithm is applied to several complex scenes with textured objects and strong interreflections. The separation results are then u sed to solve three problems pertinent to visual perception; determinin g illumination color, estimating illumination direction, and shape rec overy.