Inferring Segmented Surface Description from Stereo Data

We present an integrated approach to the derivation of scene description from binocular stereo images. Unlike popular stereo approaches, we address both the stereo correspondence problem and the surface reconstruction problem simultaneously by inferring the scene description directly from local measurements of both point and line correspondences. In order to handle the issues of noise, indistinct image features, surface discontinuities, and half occluded regions, we introduce a robust computational technique call tensor voting for the inference of scene description in terms of surfaces, junctions, and region boundaries. The methodology is grounded in two elements: tensor calculus for representation, and non-linear voting for data communication. By efficiently and effectively collecting and analyzing neighborhood information, we are able to handle the tasks of interpolation, discontinuity detection, and outlier identification simultaneously. The proposed method is non-iterative, robust to initialization and thresholding in the preprocessing stage, and the only critical free parameter is the size of the neighborhood. We illustrate the approach with results on a variety of images.