Layered Ground Floor Segementation by Computing Plane Normals from Motion Field for Vision-based Robot Navigation

초록

This paper proposes a method of detecting movable paths in local map building for visual navigation of a mobile robot. The algorithm is to detect and segment the ground floor by computing plane normals from motion fields in image sequences. A plane normal in 3D space is an effective clue to detect other static or moving objects on the ground floor and it is computed from point correspondences and plane homography. Such plane normals are combined together with iterative refinement processes based on image segmentation techniques and then allow us to detect and segment the ground floor accurately although mismatched point correspondences are detected in image sequences. The preliminary experiments demonstrate the effectiveness of the proposed method by experiments on real data.