A global localization approach based on relative position relationships of line segments

Scan matching is a method for mobile robot global localization in indoor structured environment. The disadvantages of current methods are low computational efficiency and large probability of error matching in the large-scale environment. In this paper, we proposed a new scan-matching method for global localization with the relative position relationship of complete line segments (CLS) and visible complete line segments (VCLS). The proposed method compared the relative relationship of CLS in local map and VCLS in ordinal global map to avoid excessive coordinate transformation and limits of line segments in matching between the local map and the global map. In the simulations, new method can obviously reduce computation cost in large-scale environment.