Simplification And Deformation Of Unorganized Point Cloud

Recently, modeling and rendering based on unorganized point cloud have been paid increasing attention in computer graphics community. Such technique considering points as primitives is beyond the capabilities of traditional methods in aspects of efficiency of rendering and reconstructing, enhancing huge data sets process, and optimizing.An important geometric processing of point-based technique is surface simplification. In this paper, we present a new surface variation which has a geometric meaning, and we use it as a threshold of clustering simplification. The clusters that have a long distribution of point samples can be found easily. The experiments prove that a better simplification model can be obtained by using the new surface variation. In this paper, we also propose a clustering simplification algorithm which has a weight of neighborhood point samples. This algorithm treats the surface variation and the quantity of point samples in the neighborhood as a threshold. So the user can control the threshold to be apt to split the clusters which have more point samples and a certain curvature.The deformation of point cloud is a usual component in many areas. In this paper, we describe a common deform algorithm, and then we propose an adaptive axis algorithm. This algorithm is fast and easy to solve the problem of shape distortions as a result of a wide span rotation in the deformation regions.