Massive Vehicle Point Cloud Management Based On Distributed File System

Laser radar,is now widely used with its high precision and anti-interference ability and other characteristics.The method of vehicle laser measurement obtain the three-dimensional surface information of the surrounding environment by means of vehicle speeding and vehicle-mounted lidar longitudinal scanning,which is known for its high efficiency and high precision.However,the massive three-dimensional point cloud data caused by the large-scale terrain has exceeded the current single computer storage management capabilities,and how to carry out the real-time scheduling of massive cloud data also become the current research problem.In view of this problem,combined with the widely used Hadoop distributed file system,this paper carry out a research about a distributed storage and scheduling technology for mass cloud data is developed.On the basis of research of index structure based on vehicle point cloud data,a combination of multi-level directory of the point cloud segmentation octree index storage and scheduling method is proposed,which lays a foundation for the further implementation of point cloud data distributed management.Further this paper study and research The Hadoop distributed file system in depth,and in order to overcome the problem that the original HDFS structure can not effectively support a large number of point cloud index of small file storage and scheduling in the experiment,introduce the HBase distributed database to support efficient index file distributed storage and scheduling algorithms of point cloud.On the basis of this,the C++ programming interface of HBase is implemented by thrift,and the mass distributed cloud management algorithm based on OSG rendering engine is implemented in this paper.Experimental results show that the method studied in this paper can effectively support the distributed storage and visual scheduling of mass cloudpoints.