A method and system for fully automatic construction of 3D scene based on 2D data

Abstract

The invention discloses a method and a system for realizing automatic construction of a three-dimensional scene based on two-dimensional data. The method specifically includes: reading and analyzing a variety of two-dimensional shp data based on GDAL; constructing an initial triangular network according to the acquired terrain control point and topographic landform line data; calculating the elevation values ​​of various data; realizing constraint deduction or deletion based on CGAL , the constrained triangular network is used as the terrain to construct a 3D model; under the OSG technology, the 3D model of each element is constructed for the green ground, residential area, water system, road surface, and road centerline according to the elevation values ​​of various data. The system includes: a shp data reading module, an initial triangular net building module, a data elevation value calculation module, a constrained irregular triangular net building module, and a three-dimensional model building module. Based on the open source OSG technology, the invention realizes fully automatic three-dimensional scene construction through reasonable design based on the existing two-dimensional data, saves the cost of realizing necessary functions, and has wider promotion value.

Description

technical field [0001] The invention relates to the field of three-dimensional scene construction, in particular to a three-dimensional scene automatic construction system. Background technique [0002] The related research on the technology of rapid construction of 3D scene has already been carried out at home and abroad and has been put into use. At present, the rapid construction of 3D scenes mainly includes: oblique photography model generation based on UAV technology, 3D laser scanning modeling, SLAM fast scanning modeling, deep learning scene modeling and traditional manual modeling. [0003] The above modeling methods have advantages and disadvantages in different fields. The oblique photography model is generated based on UAV technology. Under the background that UAV technology is becoming more and more mature and stable, it has great advantages for the construction of 3D scenes in a small area. It has high simulation degree and fast production process, but The cos...

AI Technical Summary

Problems solved by technology

[0003] The above modeling methods have advantages and disadvantages in different fields

Generating oblique photography models based on UAV technology, in the context of the current UAV technology becoming more and more mature and stable, has great advantages in the construction of 3D scenes in a small range, with high simulation and fast production process, but The cost is high, and it is generally only suitable for small-scale modeling. When the scope is large, the cost investment is too large. Moreover, the current domestic legislation on drones is not complete, and it is very easy to have disputes with various confidential units. limitations

Due to its high measurement accuracy, 3D laser scanning technology has been widely used in traditional building digital model construction methods, but it is difficult to be widely used in practical applications due to high engineering cost

SLAM rapid scanning modeling, deep learning scene modeling and other methods have their own shortcomings in terms of cost input, actual effect, and generation range consignment.

[0004] For the automatic construction of large-scale 3D scenes, the real-time effect of texture is generally not high, and it does not need the fine texture details similar to the oblique photography model of the drone, but the physical coordinates of ground objects are relatively demanding. High, current modeling methods are flawed

Images