Point Cloud Data Processing And 3D Modeling System Based On LiDAR

In recent years,laser point cloud data modeling has been widely used.In this paper,a modeling system is developed for the two-dimensional laser radar on the ground,which is harmless,non-contact and little error.It is applied to some irregular shapes of wood and small plants.The system can scan quickly,modeling well which has the function of calculating model volume.Firstly,this paper uses C++ language to control the sliding table,and then designs the scanning mode based on the ground lidar.Then using C#TCP/IP communication principle to make the upper computer with seanning radar,and eompleted the setting of scanning parameters and the realization of a variety of scanning methods.Then,using the Winform framework to develop the data processing interface.Converting the collected data into the coordinate values of the points needed for modeling.The main steps include de-noising and simplifying the point cloud,converting the distance value to the coordinate value and drawing the two-dimensional image of the front and back of the object.In the following modeling step,this paper uses the WPF framework to import the point cloud into the 3d window and establish the surface.Point cloud splicing is completed and a complete 3d model is formed.WPF framework takes XAML to write the interface and C#code to implement custom functions.Finally,iteration method,Delaunay triangulation method and improved splicing volume method are used.So,the point cloud data of irregular wood collected in forestry production are acquired,the three-dimensional model is established,and the volume,surface area and mass of the model are obtained.The results show that this method is closer to the reference value,and the relative error range is 0.22%?0.27%.After many experiments,the optimal working frequency,angular resolution and sliding speed are obtained to ensure the calculation accuracy of volume and surface area.It provides technical support for the three-dimensional modeling application of point cloud data in forestry and related fields.