A consistent map building method based on SURF loop closure detection

Abstract

This paper proposes a consistent map building method for mobile robot navigation, using laser range finder scans. The method is divided into three steps: Firstly, the ICP (Iterative Closest Point) algorithm is applied to match subsequent frames of laser range finder data; Secondly, the SURF (Speeded Up Robust Features) algorithm is adopted to find loop condition, which is of great benefit for correcting the accumulated error of robot pose; Finally, a global consistent map is built by Extended Kalman filer (EKF) algorithm. Experimental results indicate that the proposed method can efficiently build global consistent maps.