DIRECT DEPTH AND COLOR-BASED ENVIRONMENT MODELING FOR MOBILE ROBOT NAVIGATION

摘要

This paper describes a new method for indoor environment mapping and localization with stereo camera. For environment modeling, we directly use the depth and color information about each image pixel as visual feature. In addition, we use only the depth and color information at horizontal centerline in image, where optical axis passes through. The usefulness of this method is that we can easily build a similarity measure between modeling and sensing data on the horizontal centerline. Based on such compact environmental abstraction, we proposed a map building method about indoor environment and a localization method for estimating mobile robot positioning with random sampling stochastic algorithm. With basic real experiments, we show that the proposed method can be an effective visual navigation algorithm.