抄録
The accuracy of vision sensors and distance sensors to be used for environmental recognition is dependent on environmental condition. We propose method for generating 2.5-D map that includes information of impeditive bumps for autonomous mobile robots in outdoor environments. 2.5-D map has information of minute scale bumps which have a bad influence when a robot passes. We generate a 2.5-D map in real time with using the information of the image sensor and distance sensor while moving. Result of the experiment by the proposed method in outdoor environment shows highly precise and reproducible map generation is possible.
