Aerial Perspective Image Matching With Ground View Image Based On Patch

In recent years,with the rapid development of computer technology,Tilted image 3D modeling technology has also been rapidly developed.Now,tilted image 3D modeling has evolved from semi-automatic to fully automatic,from a single platform to integrated cloud services.However,most of the 3D modeling methods are now based on aerial perspective images,for example 3D Modeling Method Based on UAV Image,although we can get a more complete three-dimensional model,however,because of the aviation perspective,three-dimensional models lack image information from other perspectives,as a result,there are holes in observations from other perspectives,and the model's fineness and completeness are not satisfactory.In this paper,the problem of lack of completeness and poor degree of detail in the perspective of terrestrial observations for 3D models obtained from 3D modeling using only a single aviation perspective image,add ground-view images to 3D modeling,reference PMVS optimization patch method,through the establishment and optimization of patch to eliminate or reduce the image distortion due to the large difference in angle between aerial perspective image and ground perspective image,designing a specific data acquisition method to achieve a rough scale guidance between aerial perspective images and ground perspective images,finally,matching of aviation perspective images and ground perspective images is achieved through matching of patch,provides the basis for image matching to further realize integrated 3D modeling of aerial perspective images and terrestrial perspective images,in this way,you can get a complete and clear 3D model from all perspectives.The matching of aerial perspective images and ground perspective images belongs to a category of large viewing angle image matching,problems with large viewing angle images distortion,you can also use method of this article to eliminate or reduce image distortion,to achieve a successful match.This article first uses large viewing angle images to experiment,validate the effectiveness of the algorithm.And then anti matching the aerial perspective image with the ground view image,use this algorithm to experiment,compare with other matching algorithms,a detailed analysis of these algorithms in terms of accuracy,stability,etc.Experiments show that the accuracy and effectiveness of this algorithm have certain advantages.