MODEL-BASED INTERPRETATION OF STEREO IMAGERY OF TEXTURED SURFACES

We present a scheme for reliable and accurate surface reconstruction f rom stereoscopic images containing only fine texture and no stable hig h-level features. Partial shape information is used to improve surface computation: first by fitting an approximate, global, parametric mode l, and then by refining this model via local correspondence processes. This scheme eliminates the window size selection problem in existing area-based stereo correspondence schemes, These ideas are integrated i n a practical vision system that is being used by environmental scient ists to study wind erosion of bulk material such as coal ore being tra nsported in open rail cars.