Self-closure global registration for subway tunnel point clouds

Abstract

This paper presents a high-precision global registration method for subway tunnel point clouds, which mainly consists of two parts: the pair-wise registration of point clouds based on the ICP algorithm and the global registration of point clouds under the self-closure condition. Combining the features of subway tunnel, this paper analyzes the pros and cons of the control points distribution for the current point cloud registration method, and offers an improved alternative; According to the results of the section-controlled fine registration and the self-closure condition formed in the registration process, the global registration method based on its own closed condition is adopted to eliminate the accumulative errors in the section point clouds fine registration to achieve the best overall registration of all involved tunnel data. In this paper, experimental results gained from a certain section of subway tunnel in Shanghai show that the improved section-controlled registration method can improve the point cloud registration accuracy and the global registration method can reduce the error accumulation effectively.