A Bayesian Foundation for Active Stereo Vision

Sensing three-dimensional shape is a central problem in the development of robot systems for autonomous navigation and manipulation. Stereo vision is an attractive approach to this problem in several applications; however, stereo algorithms still lack reliability and generality. We address these problems by modelling the stereo depth map as a discrete random field, by formulating the matching problem in terms of Bayesian estimation, and by using this framework to develop a “bootstrap” procedure that employs fine camera motion to initialize stereo fusion. First, one camera is translatd parallel to the stereo baseline to acquire a narrowbaseline image pair; then. the depth map obtained from the narrowbaseline image pair is used to constrain matching in a “widebseline” image pair consisting of one image from each camera. T L result of our procedure is an estimate of dcpth and depth u,,certainty at each pixel in the image. This approach produces accurate depth maps reliably and efficiently, applies to indoor and outdoor domains, and extends naturally to multi-sensor systems. We demonstrate the potential of this approach by showing results obtained with scale models of difEcult, outdoor scenes.123