Large-scene airborne point cloud semantic modeling method

Abstract

The invention provides a large-scene airborne point cloud semantic modeling method. The large-scene airborne point cloud semantic modeling method specifically comprises the following steps: 1) classifying ALS point cloud scenes; (2) building semantic element identifiers; (3) building semantic element identifiers; (4) making precision evaluation. The method has the advantages of (1) fusing multiplecurrent modeling ideas to reflect flexibility of a modeling method, (2) improving the possibility of processing large-scale point cloud data through an algorithm, and (3) guaranteeing the integrity of model information in the aspects of geometry, topology and semantics.

Description

technical field [0001] The invention relates to a large-scene airborne point cloud semantic modeling method, which belongs to the technical field of architectural model measurement and construction. Background technique [0002] Extracting the geometric model of buildings is the foundation and key of building a digital smart city, and it is also the focus of current digital city construction. LiDAR technology (Light Detection And Ranging, LiDAR), especially Airborne Laser Scanning technology (Airborne Laser Scanning, ALS), as an important means of collecting large-scale three-dimensional spatial information of buildings, has the advantages of short data acquisition cycle, high precision and timeliness With the characteristics of high precision, strong initiative, and large scanning scenes, it has gradually become an important way to collect 3D architectural data in large scenes. With the development of laser payload hardware technology and storage technology, the accuracy a...

AI Technical Summary

Problems solved by technology

[0005] (1) Boundary expression modeling does not make full use of the reliable characteristics of point cloud surface information, so that the geometric accuracy of the created 3D model at key points and boundary lines is not high;

[0006] (2) Strict prior assumption modeling can improve the abstract granularity and regularity of the model, but at the same time it will sacrifice the geometric accuracy of the model, which is not conducive to the reconstruction of complex structural buildings with various shapes in large-scale point clouds;

[0007] (3) Although the dimensionality reduction modeling idea simplifies the modeling problem and improves the robustness and scalability of the algorithm, it weakens the comprehensiveness and integrity of the information expressed by the point cloud in the three-dimensional space;

[0008] (4) Divide and conquer strategy modeling often needs to recognize the overall structure of the building from the semantic scale of the parts that make up the object, so as to divide the building space reasonably, but the current algorithm is mostly based on the analysis of local geometric features, and divides the building at a lower scale , seldom take into account the overall structural expression of the building;

[0009] (5) Although nonlinear modeling improves the geometric accuracy of the model, it sacrifices the semantic expression of the model and weakens the salience of the model boundary

Images