Research On Processing And Reconstruction Of 3-D Laser Point Cloud

With the wide application of computer technology in related field, the advantage of reverse engineering revealed gradually. Reverse engineering mainly includes the point cloud data acquisition, the pre-processing of the point cloud, three-dimensional reconstruction and so on. The pre-processing and three-dimensional reconstruction of the point cloud has an important foundation role in reverse engineering.Research in this area has important significance on improving the efficiency of data processing and reconstruction.This paper focuses on pre-processing and three-dimensional surface reconstruction of point clouds in reverse engineering.Main work of this paper is as follows:(1)Point cloud data in the process of acquisition will inevitably be mixed with noises which will affect the subsequent three-dimensional reconstruction seriously. The existence of outliers will affect the normal and curvature estimation in the late,so first remove these outliers;then limit the neighborhood points to the most similar area through the vector correction of point cloud;finally trilateral filtering method is used to these noise point. This method can not only guarantee the correctness of the direction of filter, but also can keep the details of the target surface characteristics.(2)Massive point cloud data can affect efficiency and complexity of subsequent point cloud processing. In this paper,considering the accuracy?precision and speed of simplification,it uses a combination of the point cloud simplification algorithm.First,the point cloud is divided unevenly.Then simplify the point cloud by using curvature simplified algorithm in each small partition.Finally realized the simplification of the massive point cloud while maintaining the characteristics of point cloud model.(3)Point cloud reconstruction by rapid increment of triangular mesh method.First of all,concentrate the candidate points within a dynamic sphere by dynamic ball strategy to shorten the time of looking for candidate. Then find the most advantages of the boundary edge to expand based on multiple constraints principles,at the same time using the distance calculation instead of trigonometric calculations to improve the efficiency of triangular grid.Finally,select the most advantages candidate through the evaluation function.Connect this point with boundary edge to compose a triangular mesh.Finally, we analyze the experimental results and verify the feasibility and effectiveness of the proposed method.