MOBILE ROBOT NAVIGATION AND SCENE MODELING USING STEREO FISH-EYE LENSSYSTEM

We present an autonomous mobile robot navigation system using stereo f ish-eye lenses for navigation in an indoor structured environment and for generating a model of the imaged scene. The system estimates the t hree-dimensional (3D) position of significant features in the scene, a nd by estimating its relative position to the features, navigates thro ugh narrow passages and makes turns at corridor ends. Fish-eye lenses are used to provide a large field of view, which images objects close to the robot and helps in making smooth transitions in the direction o f motion. Calibration is performed for the lens-camera setup and the d istortion is corrected to obtain accurate quantitative measurements. A vision-based algorithm that uses the vanishing points of extracted se gments from a scene in a few 3D orientations provides an accurate esti mate of the robot orientation. This is used, in addition to 3D recover y via stereo correspondence, to maintain the robot motion in a purely translational path, as well as to remove the effects of any drifts fro m this path from each acquired image. Horizontal segments are used as a qualitative estimate of change in the motion direction and correspon dence of vertical seg ment provides precise 3D information about objec ts close to the robot. Assuming detected linear edges in tile scene as boundaries of planar surfaces, the 3D model of the scene is generated . The robot system is implemented and tested in a structured environme nt at our research center. Results from the robot navigation in real e nvironments are presented and discussed.