Large-scale Overlapping Mesh Optimization Algorithm

The 3D world made of lots of mesh models plays an important part in people's life. Because of more and more requirement for 3D world, lots of mesh data needs processing, storing, computing, displaying and transmitting, which brings many problems. At the same time, there are many overlaps between the mesh models. So it is necessary to optimize these overlapping mesh models.After many years of research, many scholars have come up with kinds of optimization algorithms, such as area merge algorithm, vertex clustering algorithm, wavelet decomposition algorithm, vertex culling algorithm, progressive meshes algorithm. Although these algorithms work well, they can only be used in some fields, and they are complicate, time-consuming. It is not necessary to display all meshes in 3D scene, and there are many overlapping meshes, so combining meshes is a good way to simplify models.Given these problems, a simple and efficient mesh optimization algorithm for large-scale overlapping mesh models is proposed in the thesis. This method integrates mesh merging and mesh simplification the two kinds of mesh optimization algorithms, which is more efficient and applicable. To Getting rid of the overlaps between the mesh models is a good way to reuse the mesh. So an easy and workable method based on the mesh merging is proposed. The mesh models after merging can be simplified further, so an improved simplification algorithm based on the vertex culling algorithm is proposed. The improved algorithm reduces the complexity, at the same time it can keep the original character models as soon as possible. Followings are mainly the points of innovation:First of all, the mesh merging algorithm is proposed in the thesis on the basis of analysis. The classic merging algorithm is composed of many complicated steps, the detection of overlapping area, the corrosion of mesh boundary, the production of seam line, merging meshes and smoothing mesh. The complicate mesh merging algorithm makes the massive mesh models inefficiency. The improved merging method includes detecting the overlapped area using intersected bounding box, and determining the least distance between the vertex and center of mass, merging the mesh through the triangle pair, smoothing out the merging part by means of Laplacian smoothing algorithm, and controlling the process of merging by use of a series of parameters. This method makes the process simpler and more efficient.Secondly, a comparison between the kinds of mesh simplification algorithms is made in the thesis. Comparatively speaking, the vertex culling algorithm is simple and easy to achieve. On the basis of this method a improved vertex culling algorithm is come up. The process consists of the vertex deletion based on the weight of vertex, and dividing the holes produced after vertex deletion based on the principle that the shortest is chose firstly, finally finishing the mesh optimization.The massive overlapped mesh models optimization algorithm has the advantages of mesh merging and the advantages of mesh simplification. It improves the efficiency through simplifying process. And the experiment shows that the impact of optimization is very good.