A Robust 3D Circle Fitting Method and Its Application in Tunnel Ring Detection of Subway Tunnel
3D circle fitting is widely used in point cloud processing,subway construction and industrial inspection.However,current 3D circle algorithms face challenges such as cumbersome calculation steps and the susceptibility of fitting accuracy to outliers.To solve these problems,a 3D circle fitting method which combines improved principal component analysis and Rodrigues projection was proposed in this paper.First,it was proposed that the improved principal component analysis method can eliminate the gross errors in the measurement while fitting the optimal plane of the spatial circle,this procedure was both simple and efficient,and it improved the robustness of the algorithm.Then Rodrigues projection was used to project the three-dimensional coordinates onto the fitted plane for the fitting of the plane circle,and then the fitting was projected to the three dimensions,which further improved the convenience of the algorithm.Finally,a comparative experiment was carried out between the actual measurement data of the tunnel ring in a metro station in Shenzhen and several common 3D circles fitting methods to analyze the universality of the improved algorithm and the ability of eliminating outliers.The experimental results show that the fitting accuracy of the 3D circle fitting method in this paper is reliable,which can effectively eliminate the gross error caused by the observation,improve fitting accuracy and work efficiency of the 3D circle.The average roundness is 4.8 mm in the presence of gross errors,which is nearly 50%higher than the comparison method.
metro3D circleprincipal component analysisRodrigues Transformationtunnel ring
NETL
NSTL
万方数据
