Database Management & Visualization Of Huge Range Images And Digital Images

Generally speaking, a large number of range images data will be generated in the process of collecting 3D structure information of huge building based on the 3D LIDAR scanning technology. For instance, more than 500 range images data are collected in the project of digital survey in Taihedian. Every range image data is at least in 60MB. In the end, the whole size of range images data adding other palace are always more than several hundred GB. At the same time, for the modeling and computer simulation with range images data, accurate digital images data of huge building should be collected, too. Only in Taihedian, about 2000 pieces of photo are made, the size of every piece is about 8MB. It is obvious that it's not feasible to manage so huge data in OS'file management system. How to organize and manage range images data and digital images data will be a very important task in using LIDAR technology.FILE structure and Standart Template Library technology are used in reading huge range images file, which is in order to parse known format rapidly and efficiently. After rearranging the parsed data, the rearranged blocks in Bytes are put into BLOB field in Oracle databases. A new algorithm for interactive rendering of huge range images data based on view-dependent level-of-detai(lLOD)and clipping windows-dependent exchange between main memory and Oracle database techniques is presented. In the process of reading huge range images file,range images data are partitioned into small blocks with the different size according to angle information and its MBB. Meantime, 3D R-tree spatial index is created by means of MBB in main memory. In the interactive rendering, range images data outside the clipping windows are not rendered. That is, several small blocks that their MBBs intersect the clipping windows by searching in 3D R-tree spatial index should be shown on the screen. The distance scan of center point of screen to center point of MBB controls the resolution distribution of points on the screen, so that the small blocks are able to be searched and display fast. For the storage of digital images, generally, the whole file is considered as a block or BLOB to store in Oracle directly based on I/O data streaming transmission. At one time,some indispensable information should be analyzed into attribute field to store in Oracle with the block. At last, a table, which comprises range images and digital images table pointer field,corresponding points coordinate field,spatial attitude parameter field, will be created for uniform management in range images data and digital images data. The experiment system is developed in VC 6.0 program language and OpenGL 3D graphic library. The range images and digital images data sampled from several great halls of Forbidden City are used in experiment. These data are stored in Oracle databases with Oracle Objects for OLE (OO4O). Experimental results show that our approach can not only design to allow easy access to range images and digital images data stored in Oracle databases, but also realize real-time rendering for huge datasets. Those are the grounds for the modeling and computer simulation with range images and digital images data.