Research On Uncertainty Model Of Point Cloud Object

Three-Dimensional Laser Scanning,also known as real-world replication technology,different from the traditional single-point measurement technology,3D laser scanner can quickly obtain the dense point cloud of complex scenes through long-distance,highprecision,non-contact and high-efficiency work,and the point cloud contains the 3D coordinates,intensity,reflectivity,color and other information of the scanned object surface,which provides a new technical means to build the 3D model of the object,so it is applied in 3D reconstruction,urban planning,Therefore,it is used in the fields of 3D reconstruction,urban planning,heritage protection,intelligent transportation,disaster monitoring,etc.With the development of Three-Dimensional Laser Scanning,it has become an important means to obtain spatial geographic information data,and point cloud has become an important part of spatial geographic information data,which greatly enriches the data source of spatial geographic information database.However,the accuracy evaluation of point clouds consistently faces great challenges.Compared with other spatial geographic information data sources,3D laser scanner acquires huge volume of point clouds,and the data processing process is complicated and tedious,and point clouds introduce uncertainty from the data acquisition stage,and the uncertainty will accumulate and propagate in the data processing process,which affects the quality of point cloud data and point cloud products,so it is necessary to It is necessary to evaluate the uncertainty of point clouds.In the field of point cloud uncertainty research,there is a lack of research on the uncertainty of point cloud spatial objects.Therefore,this paper proposes to use the basic spatial objects in point clouds,3D points/lines/planes,to evaluate the quality of the overall point clouds,to analyze the uncertainty of 3D points/lines/planes in point cloud space by using the related theory of uncertainty,and to derive the uncertainty model of 3D points/lines/planes.Several aspects:(1)Construction of uncertainty model for feature parameters of point cloud space objects.The 3D points/lines/planes in the point cloud are expressed using the least number of feature parameters,and the uncertainty model of the feature parameters of the point cloud space objects is derived.(2)Construction of uncertainty propagation model for point cloud spatial objects.In the context of the construction of real-world 3D China,in order to ensure consistent coordinate reference of geospatial data,the coordinate system of geospatial data is generally unified to the georeferenced coordinate system for use in the surveying field,so this paper derives the uncertainty propagation model of point clouds in the process of georeferencing.(3)Construction of uncertainty visualization model of point cloud spatial objects.By studying the uncertainty matrix of 3D points/lines/planes,the visualization index of 3D spatial objects is abstracted from the uncertainty matrix,the uncertainty of 3D spatial objects is visualized,and the color information is added to the visualized graphics.(4)The application of uncertainty of point cloud space objects.In this paper,the uncertainty applications of 3D points and 3D straight lines are introduced.In the application of 3D point uncertainty,the equation of bit difference of 3D points on any plane is derived,and 3D straight lines mainly introduce the application of 3D straight line uncertainty in the tilt deformation of buildings.