Building extraction method and device based on TIN data

Abstract

The invention provides a building extraction method and device based on TIN data and relates to the field of image processing. The method is characterized in that accurate datum points are determined, appointed points close to the roof of a target building are selected as the datum points, and the problem that the acquired contour points of an independent building are inaccurate due to the blocking of trees or other barriers is avoided; three-dimensional points, higher than the datum points, in three-dimensional space are used as the reference points of the target building; due to the fact that the edge points of the three-dimensional points are used as the contour points, the acquiring range of building data is determined according to the outline formed by the contour points, and the three-dimensional model of the target building is extracted.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a method and device for extracting buildings based on TIN data. Background technique [0002] TIN is an irregular triangular network, also known as "surface data structure", which is a kind of vector data. The area is divided into equal triangular surface networks according to the finite set of points in the area. The digital elevation model is composed of continuous triangular surfaces. The shape and size of the triangular surfaces depend on the density and location of irregularly distributed measuring points. [0003] Continuous 3D texture TIN data is continuous integrated stereo model data generated based on multi-view images. It is different from 2.5D data that lacks side information. It is true 3D data that can represent convex and concave surfaces. The data is composed of 3D TIN and texture images. , can truly express the surface information of the surface and objects, and ...

AI Technical Summary

Problems solved by technology

[0004] However, most of the 3D texture TIN data generated by existing software is a continuous integrated three-dimensional model, and the ground surface and ground objects are connected together, which cannot analyze and study a single entity object (such as a building), which severely limits the ability of different entities. application of elements

The building model is the main entity element in the ground object, so it is necessary to extract the building model from the continuous and integrated 3D texture TIN data for editing, attribute attachment, hierarchical classification management, etc., to meet the requirements of physical management and However, it is a technical difficulty to maintain the seamless connection between the extracted building model, terrain data and texture, and the level and structure of the continuous integrated three-dimensional model, which directly makes it difficult to extract Standalone building models, or cause extract model to fail

[0005] At present, there is no building extraction method based on 3D texture TIN data

Images