Maintaining representations of the environment of a mobile robot

A description is given of current ideas related to the problem of building and updating three-dimensional representations of the environment of a mobile robot that uses passive vision as its main sensory modality. The authors attempt to represent both geometry and uncertainty. The authors motivate their approach by defining the problems they are trying to solve and then give some simple didactic examples. They then present a tool they think is extremely well adapted to solving most of these problems: the extended Kalman filter (EKF). The authors discuss the notions of minimal geometric representations for three-dimensional lines, planes, and rigid motions. They show how the EKF and the representations can be combined to provide solutions for some of the problems. A number of experimental results on real data are given