Research On Extraction And Visualization Of Slope Terrain Based On Laser Point Cloud

The slope of the mountainous expressway often collapses and landslides due to natural or artificial factors.It has a negative impact on the road traffic in the surrounding areas,the people’s life and the economic income.In order to predict and prevent similar accidents,it is necessary to survey the slope terrain along the highway.Deformation trends of different slope terrain are calculated through data comparison and analysis.In many methods of topographic monitoring,three-dimensional laser scanning has the unparalleled advantages to other methods.First,laser point cloud can directly measure the three-dimensional coordinates of the surface of the object without the matching of same-name points and the solution of collinear equation.Secondly,laser can penetrate the vegetation partially and produce two echoes or multiple echoes,which is beneficial to the extraction of the real surface of the terrain.Finally,lidar mounted on a small UAV with other location devices can form airborne lidar.The degree of automation is greatly improved,and more target areas can be measured in a short time.In this paper,the laser point cloud is used to reconstruct the slope terrain,and the terrain visualization system is realized,studing the realization method of 3D laser scanning by 2D laser radar then mathematical model of coordinate transformation of the LIDAR system has been established.This paper proves that RPLIDAR A2 can satisfy the point cloud measurement need.PC software platform to drive laser radar sensor and sensor’s base has been developed.Then the article discusses the process of point cloud processing.The method of point cloud registration is introduced to splice the whole point cloud of the target area.Three methods of point cloud filtering are put forward.After comparing and analyzing,it is concluded that the iterative weighted algorithm based on least squares and the moving fragment discriminant method are suitable for slowly-changing slope terrain covered by sparse vegetation,and the filtering effect is good.Next,point cloud simplification model is studied,and the neighborhood point cloud simplification method based on feature points is elaborated,which can sharply reduce the number of point cloud and preserve the geometric feature points.It improve the speed of point cloud processing greatly.In the end,the method of building digital elevation model based on irregular TIN mesh and Grid mesh is described as well as an interpolation method for establishing TIN network and Grid mesh.Terrain visualization software has been developed by OpenGL under the MFC framework.