Ground height map generation for a humanoid robot in an unstructured environment

This paper proposes a method to measure an accurate ground height map for a humanoid robot by using a laser scanner in an unstructured environment. To measure accurate dimensions of a surface or an obstacle, a 2D laser scanner is installed on the head of a humanoid robot, called MAHRU-1. By moving the head up and down, 3D depth map around a robot is obtained from the 2D range data. After obtaining the 3D depth map around a robot, a plane segmentation method is applied to the 3D data. At first, all the horizontal and vertical planes in the 3D point map are extracted, and the outliers are removed from the plane. After segmenting the plane from the 3D data, the dimensions and the area of the planes are calculated. The accuracy of the extracted surface is evaluated with experimental results, which show the effectiveness of proposed method to extract the surfaces in the unstructured environment.