Specularity removal for shape-from-shading

Specularity reflecting surfaces confuse traditional shape-from-shading algorithms because the variation in image intensity within a specularity does not directly relate to the cosine of the incident angle, as it would for a simple Lambertian reflector. To overcome this problem. color is introduced and a method of removing the specular component of the intensity variation is proposed based on a dichromatic model of surface reflection. Unlike Shafer's method for specularity removal, which is restricted to uniformly colored surface patches, our algorithm uses information from several differently colored regions. The problem of segmenting an image into color regions is successfully avoided as the specular component is calculated and removed using local operations only. The image resulting from specularity removal preserves the relative intensity of the diffuse component so it can then be input to a shape-from-shading algorithm. Our shape-from-shading algorithm is based on variational calculus. Without assuming the location of the scene illurninant, and allowing background illumination, the algorithm computes the shape from the diffuse component image in a more general setting than the existing algorithms do. In the thesis, the algorithm is formulated into a local relaxation scheme which allows a parallel network implementation.