An Uncertainty Model of Stereo Vision and its Application to Vision-Motion Planning of Robot

Abstract

This paper describes an uncertainty model of stereo vision and its application to a visionmotion planning for a mobile robot. In general, recognition of an environment requires much computation and the recognition result includes uncertainty. In planning, therefore, a trade-off must be considered between the cost of visual recognition and the effect of information obtained by recognition. Such a trade-off must be formulated on the basis of a model of vision which describes the required time for visual processing and uncertainty of information to be obtained . In this paper, an uncertainty model of stereo vision is described, in which not only the quantization error but also false matchings of features are considered. A strategy for resolving ambiguous matchings is also proposed . Using the uncertainty model in the planner , an optimal plan for a real world problem is generated . An efficient solving strategy is also described which employs a pruning method based on the lower bound of the total cost calculated by the assumption of perfect sensor information .