Straight Skeleton Based Automatic Generation of Hierarchical Topological Map in Indoor Environment

Real-time autonomous navigation and scheduling are regarded as the prerequisite capabilities for indoor mobile robots, while the application of prior information about the indoor map can contribute to reducing the level of computational complexity significantly. The indoor maps can be generated through manual marking or using automatic generation methods. The latter considerably reduces human intervention, which makes it more suitable for large indoor environments. However, most of the existing approaches to the automatic generation of indoor maps suffer such problems as imprecision and the lack of smoothness. These issues impede the generated map from meeting the requirement for safety and traceability. In order to address these issues, this paper proposes a novel approach to the automatic generation of the indoor hierarchical topological map (HTM) based on the straight skeleton. This method involves three steps. Firstly, the boundaries of walls and obstacles are extracted from the grid map or the CAD drawing. Secondly, the extracted boundaries are used to automatically generate the primary tracks in narrow corridors, the regions in large open areas, and the secondary tracks within these regions. Thirdly, for the robots that are incapable of taking a spin turn, the clothoid-based smoothing method is applied to ensure the curvature continuity of generated tracks. According to the experimental results, the proposed method is accurate in describing the topological structure of complex and changeable indoor environments and effective in providing smooth and easy-to-track trajectories.