抄録
Map information is important for path planning and self-localization when mobile robots accomplish autonomous tasks. In unknown environments, mobile robots should generate an environment map by themselves. Then, we propose a method for construction of the 3D environment model by a mobile robot. The proposed method is based on structure from motion. We use a single omni-directional camera to measure the objects in the environment. The propose method estimates the camera motion with a feature flow model. The feature flow model makes it possible to estimate the camera motion precisely by taking advantage of the wide field of view of the omni-directional camera. Experimental results showed the effectiveness of the proposed method.
