3D Visualization Volume Measurement System’s Research And Realization

With the rapid development of science and technology, more and more data was generated by scientific computing and observed experimentation. Meanwhile, the rapid increase in computer power made that using computer graphics and image processing technology to display the intermediate and final results of scientific computing visually and directly become possible.3D visualization volume measurement was generated in this context. It has a wide range of applications, such as virtual reality, heritage conservation, mechanical processing, special effects production of film, computer simulation, fashion design and so on.The current mainstream technologies of3D visualization volume measurement are analyzed comprehensively. The three major steps, collection of scattered points, reconstruction of3D model and volume measurement which make up of those technologies are being studied thoroughly. Because of inefficient calculation based on scattered points without topology by current algorithms, a rapid volume calculation method is proposed. First, pretreat the obtained scattered point cloud, like removal of noise points, point cloud simplification, etc; Second, using the improved Delaunay Triangulation to build tetrahedral mesh; Third, calculating the fitting surface and minimum spanning tree to get normal vector of each point based on k nearest neighbors; Fourth, removing vitro tetrahedron by the point vector; Finally, calculating volume of each tetrahedron to get total volume. The method’s major improvement is build tetrahedral mesh by Delaunay Triangulation to calculate object’s volume instead of time-consuming surface reconstruction compared with traditional method.Finally, a3D visualization volume measurement system is implemented, and the rapid volume calculation method above is applied to the system. The system is built based on CGAL. The main features are point set processing, surface reconstruction, mesh processing and volume calculation. Then, the algorithm above is verified by the system using several scattered point clouds, and is compared with traditional algorithms from two aspects of efficiency and accuracy. The results show that the rapid algorithm not only improves the computation efficiency greatly and ensures certain accuracy.