Mesh And Surface Reconstruction Processing For The Car Seat Point Cloud Data

With the rapid development of the automobile industry, the design and produce of the car model is increasing, car seat as the important part of the car contact the driving personnel all the time, so it is especially in the design. Based on the existence excellent physical model, processing and modifications it. And how to completely and accurately digitized the excellent model and input them to the computer is a complex task. Reverse engineering is the general term of these processes. Currently, extraction of point cloud data in reverse engineering mostly use the non-contact laser scanning, extract the points layer by layer. Triangulation and surface reconstruction of point cloud data are several important issues in reverse engineering. In this paper, studied the above problems. Based on VS2010development platform, combined with the MFC and OpenGL library with their consistency and powerful feature, algorithm-based, developed a friendly interface, simple, practical car seat point cloud data processing software.The main research work is that with the characteristics of stratification, preprocessing the cloud data layer by layer. Store the cloud data into a data structure which is similar to adjacent table. Designed and implemented an improved regional growth triangle method according to the regional growth triangle method, reduce points those were selected, put forward that point and linear spatial relationship to exclude illegal triangle. The triangulation method in this paper is number the point data by the order in file. Establish a set of points, a set of edges and a set of triangle. Start from the initial generated triangle, select element in triangles which has been constructed. The best point is selected from layers those are same or near with the points on the edge. Then delete the points from the original collection and use z value to reduce the selection range. This paper’s time complexity is faster than others triangulation methods with the stratification cloud data. And have better experimental results, the mesh is clear and fast. While introduce the concept of triangulation, analysis the feature and time complexity of the Delaunay triangulation, projection method, regional growth, alpha-shapes triangulation and improved regional growth triangle method.Studied several triangular mesh storage structures those used to store triangular mesh after built. For later display, modify operation. In observation mode, operate with rotate, zoom and translation. In modify mode, point, triangle is the basic unit of the model. In this paper, use simple mesh storage structure. Use simple mesh storage structure and OpenGL naming mechanism and picking mechanism to modify the mesh. The operation is simple, fast, picked precisely. Finally, study the surface reconstruction method of the point cloud data. Use the method based on the line partition surfaces. Introduce the Bezier surface reconstruction method, B-spline surface reconstruction method, and their advantages and disadvantages. Based on the tensor product surface’s anti process, inverse calculation of the cross-section curves and control curves directions by NURBS curves inverse calculation to calculate the control points and node vector. Then draw the surfaces. Finally, smooth stitching all surfaces.