The Research And Application Of3D Laser Scanning Techniques In Land Subsidence Monitoring

3d laser scanning technology offers a brand new approach in getting spatial data.The3dlaser scanner scans surface of an object and get its3d coordinates, which will be processedlater. Finally we get a3d model of the object.3D laser scanning technology serves as a majorapproach in getting spatial data after GPS. This thesis introduces the technology into surfacedeformation monitoring presented by ground subsidence. In this way we can set up ahigh-density deformation monitoring network above the curved surface and thus get morecomprehensive data.The thesis introduces the basic principle, classification, features and major fields of useconcerning3D laser scanning technology. With the example of Leica ScanStation2, the thesisalso elaborates the procedures, notes and pretreated content in the collecting point cloud datawith3D laser scanner. The thesis highlights the method of using bi-cubic interpolationmethod to curve fit the point cloud data. Then the thesis points out that by using waveletneural network (WNN) method for surface reconstruction, and by using trained network toconduct nonlinear approximation for point cloud data, we can get a implicit surfaces of pointcloud data. At the same time also with the traditional BP neural network to surfacereconstruction of point cloud data, by comparing the two kinds of the fitting precision ofnetwork model, the wavelet neural network is obtained the conclusion of high fitting precision.In the end the thesis processes the experimental data. Based on MATLAB platform, the thesiscompletes surface fitting for point cloud data with the method of WNN and successfullyextracts ground deformation information.