Research On Index And Compression Method Of Laser Point Cloud Data

Three-dimensional laser scanning technology is an advanced automatic high precision stereoscopic scanning technology.High precision 3D model has been widely used in all walks of life,so it is very important to study the method of point cloud data processing.On the basis of the principle of 3D laser scanning technology and the research of point cloud data processing technology,this paper improves the classical algorithm of point cloud data index and point cloud data compression,in order to obtain efficient query efficiency of point cloud data and high precision point cloud data compression quality.Point cloud data index mainly include grid index,KD tree index,R tree index,octree index,four forked tree index and so on.Among them,four fork tree index has good index efficiency,but there are many problems of tree depth,four forked tree storage redundancy and stack overflow in the index process.Therefore,this paper improves the four fork tree index to improve efficiency and efficiency.First,the point cloud data is partitioned according to the span of the point cloud data,then the point cloud data after the block is indexed by the four fork tree.In the four forked tree index,the concept of the custom stack and the smallest outsourced rectangle is introduced,so as to improve the index structure of the four fork tree of the point cloud data,through the traditional four fork tree knot.Compared with the experiments,the improved point cloud data index has good tree building efficiency and query efficiency.The method of point cloud data compression mainly includes the method of curvature sampling,random sampling,uniform grid sampling,coordinate increment and regional center of gravity,in which the region barycenter method has better compression accuracy,but the surface details of some objects will be lost in the process of compression.Therefore,the region barycenter method is carried out in this paper.It improves the quality of compressed point cloud data and improves the accuracy of building 3D models.First,the bounding box is constructed on the point cloud data,and then the bounding box is divided into a number of sub bounding boxes according to the partition threshold.Then the point cloud data in the sub bounding box is replaced by the center of gravity,then the remaining points are deleted from the threshold to the nearest neighbor plane,and then the reserved points are two from the point to the center of gravity.After deleting and finally completing point cloud data compression,the improved algorithm improves the compression quality,ensures the data reduction and improves the accuracy of the 3D model by comparing with the regional center sampling and other compression algorithms.