Design And Development Of 3D-Visualization Mine System Based On GLScene

In recent years, a variety of mining accidents have occurred, particularly serious accidents have frequently appeared in newspapers. The mine safety management situation is becoming increasingly severe. In the event of a mining accident, it's often difficult to grasp the dynamics of the underground distribution of the staff, and rescue the person under mine on the lack of reliable information. Emergency rescue goes in a low efficiency, and the success rate is small. The 3D-Visualization Mine System Based on GLScene introduced in this thesis, draws the real 3D-Visualization Mine Scene according to the true scale of the mine. Through the integration of information of mine entity and mine tunnel, the laneway in the 3D-Visualization scene can carry the required information. The system gives a real-time tracking show combined with staff positioning algorithm. All of this lays a foundation for information in the mine disaster prevention.This thesis applies MVC to design the 3D-Visualization Mine System based on GLScene, and designs the model, view and controller to avoid the problems which because of the procedure structure's unreason existed in traditional software development, and achieves a loose coupling idea. At the same time, in order to optimize the system's performance, access layer based on connection pooling and JSON data communications has been designed.This thesis focus on the analysis and development of the 3D-Scene's visibility calculation and it's spatial index. Efficient visibility determination can greatly save computation on the invisible parts, improve the performance of rendering scenes. The thesis introduces a spatial index structure named QR-tree based on quad-tree and R-tree. The research indicates that QR-tree always demands more storage than R-tree, but it gains better performance in insertion, deletion, and especially searching and more the amount of spatial data is, the closer the storage cost of QR-tree and R-tree are, the better performance QR-tree has. That is to say, for very large spatial databases, QR-tree processes more superiority than R-tree. The QR-tree index improves search performance and speed up the response.