A Research On Surface Information Processing And Rendering Of Point Cloud

With the rapid development of 3D scanning technology, Point cloud has become one of the most popular 3D modeling techniques. For point cloud models, the data is easier to acquire and its data structure is much simpler than traditional mesh models. With these advantages, point cloud is very good at representing extremely large or complex models, such as sculpture and other irregular shaped objects. The surface information of a point cloud model is very abundant. It can include position, normal, size, color, texture and transparency etc. This paper is mainly about the research on the rendering of point cloud model, and the editing of point cloud surface information.After a thorough study of the currently prevailing rendering algorithms, we carefully analyze the disadvantage of multiple-pass rendering algorithms. We proposed a new single-pass rendering algorithm which overcomes the disadvantage of multiple-pass rendering techniques. The algorithm is efficient and the resulting image has high quality. Our algorithm is robust and is easy to implement.Parameterization is a fundamental tool for the editing of point cloud surface. After applying parameterization to a point cloud surface, we can convert 3D points into a parameterization domain, typically a plane or a unit sphere. After that, we can perform various kinds of editing operations on the parameterized points which will then be mapped to points of the point cloud model. This paper introduces a few parameterization techniques, we then apply them to texture mapping on point cloud surface.Point cloud morphing is a topic about how to transfer a source model into a target model smoothly. Because there is no topological information between the points, it is more difficult to implement point cloud morphing than morphing with mesh models. In this paper, we analyze previous work on point cloud morphing and compare the advantages and disadvantages of the various algorithms. We then introduce spherical parameterization into point cloud morphing, as a result, our algorithm achieves a good morphing effect for genus-0 models.