This study proposes an innovative solution integrating laser scanning and virtual engines to capture pipeline point cloud data. Using point cloud feature extraction and instance segmentation, a structured differential fitting approach tailored to pipeline types enables precise geometric parameter extraction. An automated modeling plugin developed via the Revit API converts point cloud data into BIM models efficiently. Experimental validation on a Shanghai sports stadium's first-floor pipeline system demonstrates over 80% efficiency improvement while maintaining LOD 300 accuracy. This research offers a precise and efficient solution for intelligent modeling and operation-maintenance of MEP systems, with broader applications in complex pipeline engineering.