Mesh Simplification For Huge Mesh Dataset

3D reconstruction of medical slices is a new technology which rapidly developed last decade. It mainly deals with the sequential images of biologic organizations and obtains the 3D images to do some research.With the development of science and technology, the meshes reconstructed and used in the field of medical images are more and more complex. These complex meshes are often formed of millions of triangles.Sometimes such high-resolution meshes are not needed. And the meshes' fine level and process time should be balanced. So the original meshes with large number of triangles could be replaced with the simple ones.Simplification means to reduce the number of triangles, edges and vertices of the meshes and be close to the original ones in appearance as much as possible. It is significant for the meshes'storage, transmission, and especially, real-time rendering. At the same time people's demand on the speed of 3D graphics processing are higher and higher. The traditional solution is to use high-end multiprocessor and multi-pipeline graphics workstations to finish the task. However, the triangles of slice dataset can be concurrently simplified through dividing the mesh into some continuous parts and then sending the parts to different computing nodes to perform simplification operation. This procedure will lose the topological information of the mesh at the joint between two continuous parts which is necessary for subsequent mesh operations. In this paper, we design a new file format to store the mesh data and develop some algorithms based on this format. The result shows that it can keep the topological information very well and consequently, provide the basic data for distributed parallel simplification of segment mesh data.