Feature Extraction and Detection of Intersecting Surface Features in Large TKY Pipe Joint Grouping Based on the Ransac Algorithm

In response to the challenge of detecting interfacing surface types during the automated grouping process of large conduit frames, this study proposes a quantitative evaluation model based on a certain confidence level for detecting defects in the interfacing surface. Initially, the point cloud of the branch pipe interfacing surface is segmented into multiple layers to reduce the nonlinear components of each layer. Subsequently, the RANSAC algorithm is employed to extract the largest linear component from each layer, based on which the center position of the corresponding main conduit for that layer is estimated. Finally, an approximate confidence interval is provided by discussing the probability distribution of the estimated center positions. This interval determines whether the branch pipe interfacing surface meets the assembly requirements at that particular layer. Simulation results confirm the effectiveness of this method.