A Distributed Simplification Method with PC Cluster

Abstract

Simplification algorithm of 3D model is basic, important technology that can reduce the redundancy of the expanding 3D shape data, and that can apply to various applications. The QEM method is given as an algorithm to simplify data while preserving the feature of 3D model. The QEM method is one of the algorithms with the best balance between details and the simplification speed after the vertex is degenerated. However, because the QEM method is a successive algorithm that can delete only one edge at a time, it takes time for simplification of the huge 3D models. PC cluster is one of the environment that can enlarge computational resources with uniting many PCs. It is constructed easily and used generally. In this paper, it proposes the method made parallel by extending the QEM method, and using the PC cluster. With the PC cluster constructed by typical windows PCs, simplification high-speed by this method becomes possible while maintaining the feature of 3D model.

An efficient simplification method for point cloud based on salient regions detection

“…First, note that our method has many advantages compared to some related work where the user can control the simplification rate. Furthermore, there are some cases where the simplification rate is very high [37] and some of sharp region points that conserve the characteristics of the surface shape will be removed resulting holes in the simplified 3D point clouds. In contrast, our method preserve the characteristics of the surface shape with a minimal configuration (x = 0, ∝ = 0) due to the technique we use where we maintain internal and boundary feature points.…”

“…In contrast, Figures 12 and 13 shows that our method produce better results with models of different shapes (symmetric and no symmetric). Table 3 shows a comparison of some results from [18,37] and our method. One can see that our method gives simplified models with lesser points than other methods and maintain the boundaries and ridges of 3D models.…”

Feature-Preserving Simplification of Point Cloud by Using Clustering Approach Based on Mean Curvature

“…The algorithms are roughly divided into two categories: mesh-based methods and point-based ones. Simplification based on polygon meshes requires reconstruction of triangular meshes from a point cloud and then redundant points are removed [5,6,7]. In contrast, simplification based on points does not require reconstruction of a mesh model and it only relies on the information of points to simplify a point cloud.…”