Landslide Deformation Analysis And Prediction Of Ground 3D Laser Point Cloud Data

Landslide is a kind of natural disaster that have complex structure and strong uncertainty factors.The non-contact terrestrial three-dimensional laser scanning technology(TLS)can obtain high-precision and high-density point cloud monitoring data quickly,and can let the whole deformation of landslide body conduct scientific analysis and reasonable prediction.This article make a detailed classification of TLS,application areas of TLS are discussed and summarized,and make a detailed theoretical research about the processing flow of 3D laser point cloud data.Adopt landslide point cloud data of Ludian earthquake disaster area,use Goemagic Studio 12 to carry out data processing and build model,two different periods of landslide monitoring data model are compared from 3D to 3D.Geomagic Qualify 12 is used to analyze the 3D deviation on the surface of the two landslides model.The "deformation-color" deviation profile is established for the three-dimensional displacement component.The "deformation-color" deviation distribution of the three-dimensional displacement component is established,and extracted the feature points and section line and make a comparative analysis about the"displacement deformation".Used the Sufer 7 to perform the comparison and analysis about the original point cloud,and use "Cloud Compare" to implement ICP registration and local matching for two point clouds,and perform the Haosdorff Distance comparison;The volume of the two-point cloud is calculated and compared with the volume change of the IMInspect module in PolyWorks.The conclusion is that the deformation of the landslide is not obvious and the landslide is stable.Finally,GM(1,1)gray model and GM-BP gray neural network model are used to predict the GNSS monitoring data of the landslide area.By comparing the actual deformation and the predicted deformation,get conclusion:The GM(1,1)gray model has a good effect on the prediction of landslide deformation,and the GM-BP gray neural network model has higher accuracy for landslide deformation prediction.