基于BP神经网络法的地表变形监测

Surface Deformation Monitoring Based on BP Neural Network

利用三维激光扫描技术监测地表变形时,需要对大面积的地形点云数据重建地面曲面。针对点云数据的海量性,提出利用BP神经网络的方法进行曲面重建,分别模拟出两期点云数据的曲面及两期点云数据的下沉曲面。实验结果表明,该方法对海量数据的曲面重建精度较高,并能提取变形信息,具有较高的使用价值。

We need redevelopment the surface of the large area＇s point cloud data , when monitoring surface deformation using 3D laser scanning technology .BP neural network method is introduced for surface reconstruction aimed the massive point cloud data .This method can reconstruct the surfaces of the point cloud data and elevation variations scheme of the data in different period respectively . The experiment indicated that the reconstruction of the massive data based on this method has higher accuracy and important value .