Based on 3D laser scanning technology in railway tank car and tank container (the railway tank car (container) for short) application, the point cloud which was incomplete and noise which was found during the scanning process were analyzed. Based on the massive scanning point cloud of railway tank car(container), proposed a fast and effective SG-t point cloud optimization processing method. The SG-t method included sp-H point cloud pre-processing method and Eti-G model reconstruction. The tests showed that the new methods could optimize point cloud which was noise and incomplete in a relatively short time. It could reconstruct model fast and efficient. It could greatly improve the efficiency and precision of scanning. The results which Compared with the results of capacity comparison method showed that measurement uncertainty increased from 4×10<sup>-3</sup>,k=2 to 3×10<sup>-3</sup>,k=2.Optimization and processing method of the point cloud of the 3D laser scanning of railway tank car (container) provide a reference to the development of related technologies.