Skip to Main Content
With regard to the analysis of the cloud data in reverse engineering, an effective method for data smoothing and completing is presented. This method consists of four steps: firstly the spatial unorganized point cloud should be projected to a plane, then a highly effective method called Cline-Renka is used to triangulate it. When the plane triangulation has been finished, the result should be projected to space to get three dimensional triangulation net; Secondly find the center point in each triangulation area, then take it as zero point to set up local coordinate and use a normal distribution model in the theory of probability to delete the noise points partly; After completion of above steps, the B-B surface is constructed on each triangulation net to fair data globally; Finally, in allusion to the phenomenon that partial data can’t be measured in actual data surveys, a data-completing method is proposed. Estimate the geometry information of the acquired data points which are around the unknown data points. Then set up a optimized energy model on the basis of aforementioned geometry information to complete data. Through above disposals, the point cloud data can commendably satisfy desires of the following curve and surface reconstructions.