Research And Implementation Of Three-dimensional Point Cloud Processing And Implicit Surfaces Reconstruction Techniques

In recent years, reverse engineering technology has been widely applied in the field ofmodern design and manufacturing, virtual reality, scenario modeling, computer vision,medicine and so on, it has very broad prospects. Point cloud processing and surfacereconstruction is a key content in reverse engineering. With the wide application ofthree-dimensional laser scanner, how to process large-scale and densely distributed pointcloud and how to reconstruct the surface with scattered point cloud has become a hot issue ofthe current research. The main contents of this article focused on simplifying scattered pointcloud, surface reconstruction and blending of implicit surface smoothly. Details are asfollows:This paper presents a new algorithm for simplifying point cloud. Firstly, the maximumangle between the average normal vector and the normal vector of each point in the boundingboxes should be calculated, which decides whether the bounding box need to be subdivided ornot by using segmentation method continues octree. Then the second parametric surfaceapproximation method is used to fit point cloud and calculate the principal curvatures of eachpoint which is used with the hausdorff distance to extract and retain the feature points.Experimental results show that the simplification effect of this new method is very obvious,and it can efficiently remove a large amount of redundant data and reserve the geometriccharacteristics of point cloud at the same time.Secondly, this paper presents a new implicit surface reconstruction algorithm based onthe radial basis function with ellipsoidal constraints. This algorithm uses the idea thatinteraction of two implicit surface to achieve a constraint on the radial basis function implicit surface reconstruction. Experiments show that this algorithm can fit point cloud smoothly andwell. But if there are sharp features information on the point cloud, there will be swellingphenomenon.Finally, to solve the defect of radial basis function implicit surface fitting with ellipsoidalconstraints, a new method is presented. In the meantime, a new implicit surface blendingalgorithm is proposed. The point cloud is processed with cell-based adaptive decompositionfirstly, then fit the decomposed points with the implicit surface reconstruction method basedon radial basis function with ellipsoidal constraints. At last we use a blending algorithm whichthis paper presents to blend the local surfaces for producing an overall model. Experimentalresults show that this method has many advantages of high efficiency of fitting point cloud,good characteristic of feature preserving and so on.