Exploring spherical image properties for robot navigation

This paper addresses the problem of performing navigation with a mobile robot using active vision exploring the image sphere properties in a stereo divergent configuration. The navigational process is supported by the control of the robot’s steering and forward movements j u s t using visual information as feedback. The steering control solution is based o n the diflerence between signals of visual mot ion f low computed in images o n different positions of a virtual image sphere. The majority of solutions based o n mot ion f l ow and proposed until now, where usually very unstable because they normally compute other parameters from the motion $ow. In our case the control is based directly o n the dzference between mot ion flow signals o n diferent images. Those multiple images are obtained by small mirrors, that simulates cameras positioned in diflerent positions o n the image sphere. At this momen t at is under development another new version f o r the spherical sensor. T h e control algorithm described in this work is based o n discrete-event approach to generate a controlling feedback signal for navigation of a n autonomous robot with a n active vision sys tem as described o n [2].