CONSTRAINING OBJECT FEATURES USING A POLARIZATION REFLECTANCE MODEL

A growing trend in computer vision has been the use of physical reflectance models, predicting reflected radiant intensity and color, to obtain constraints on object features. Until recently, relatively little attention has been paid to analysis of the reflected polarization state of light and what feature constraints this might provide to an automated vision system. Even though there is no analogy with human vision, we demonstrate that a wealth of constraint information can be obtained by resolving polarization components of reflected light with a polarizing filter placed in front of a camera sensor. We present a polarization reflectance model known as the Fresnel reflectance model because of its use of the Fresnel reflection coefficients. This reflectance model accurately predicts the magnitudes of polarization components of reflected light, and all the polarization-based methods presented in this paper follow from this model. We demonstrate the capability of polarization-based methods to segment material surfaces according to varying levels of relative electrical conductivity, in particular distinguishing dielectrics, which are nonconducting, and metals, which are highly conductive. Polarization-based methods can provide cues for distinguishing different intensity-edge types arising from intrinsic light-dark or color variations, intensity edges caused by specularities, and intensity edges caused by occluding contours where the viewing direction becomes nearly orthogonal to surface normals. Analysis of reflected polarization components is also shown to enable the separation of diffuse and specular components of reflection, unobscuring intrinsic surface detail saturated by specular glare. Finally, we address polarization-based methods used for constraining surface normals.