| With unique advantages of high speed,high precision and no contact by used 3D laser scanner technology.Currently,it has been widely used in many fields such as mapping,structure,cultural relics protection,deformation monitoring and so on,which has high practical application value.But,after using 3D laser scanner to obtain points cloud data and will produced a large amount of data,it bring technical difficulties of efficient organization and management of these data,evident especially in terms of search points cloud,this is because when the researchers needs so voluminous cloud of points in the search to the sense of region of interest or object is like looking for a needle in a haystack,is very difficult,or even impossible.Therefore,in order to facilitate rapid search to research needs object or region of the point cloud data,after obtaining points cloud data,using Rigel company types of Rigel VZ-1000 3D laser scanner,supporting the RiSCAN Pro V 1.8.1 of software of acquisition point cloud data of point cloud to noise and registration and other related pre-processing,qualified to handle the point cloud data to create an index is very necessary.This paper in summary on the basis of the pros and cons of point cloud tree data structure related research.The main research work is as follows:(1)Summary and analysis of advantage and disadvantage of the classical tree research status,then proposed using octree and 3D R *-tree their strengths,build a using octree and 3D R *-tree integration point cloud query scheme.(2)For the promble of the single tree structure is not harmonious with the construction and the search efficiency,this article proposes a hybrid model which uses the two.(3)In view of the problem of multi path effect of R*-tree in point cloud search,the method of kernel function and node overlap is proposed,and the optimization of node split of the two stage R*-tree structure is optimized,and the performance of the integrated tree is improved to some extent. |
PDF Full Text Request