Font Size: a A A

Research On 3D Mesh Algorithm In UAV Images Matching Point Cloud

Posted on:2017-04-25Degree:MasterType:Thesis
Country:ChinaCandidate:Q S ChenFull Text:PDF
GTID:2180330503979208Subject:Geography
Abstract/Summary:PDF Full Text Request
In recent years,more conveniently and technology integration of UAV remote sensing technology comes naturally. UAV Oblique photography can both get the integrity images and the facade textures of buildings. It is the mainstream trend of 3D modeling of urban landscape. For the large UAV image(the Sloping Image of UAV), the stitching matching technology is more mature.But, the research for Three dimensional model of the Sloping Image is less relatively,how to build three dimensional visualization model fast and accurately according to the characteristics of UAV images,it is very important problem of the application of UAV images, and it is also the one of the hot spots in the research of UAV images.Three dimensional reconstruction is the key technologies of 3D reconstruction based on 2D images,and the three dimensional construction by triangular mesh is also the most important process of the three dimensional reconstruction.For the UAV images, the three-dimensional structure by the triangular mesh is based on the existing point cloud data.Due to the three-dimensional spatial triangle mesh can not be applied to the traditional projection construct by Delaunay triangular. As a result, to find 3D triangulation method to fit UAV images is very necessary.In this paper,we presents a 3D construction method for UAV image matching point cloud data according to it is large data and rich information features 。 First of all, we present a the research background, research status and the method of 3D mesh construction are described in detail,and introduce the simplification of the triangulation mesh.Second,this article use three algorithms:MC,Poisson algorithm and Graph cuts to reconstruction the 3D triangular mesh and compared their advantages and disadvantages,on the basis of this study, using an improved graph cut algorithm and the successful reconstruction of the Large scale data is obtained.Last, simplification mesh in three different ways,comparative analysis of three kinds of simplified methods and get a better 3D model.The innovations ofthis paper are as follows:(1)Triangular mesh reconstruction algorithm based on graph cuts is proposed, which can be applied to UAV images.To provide some help for the further work of three-dimensional reconstruction of UAV image.(2)Use three different methods to simplification mesh.The applicability of the constructed 3D model is improved.
Keywords/Search Tags:UAV tilt image, 3D mesh algorithm, Matching point cloud, Graph cut, triangular mesh simplification
PDF Full Text Request
Related items