Research On Key Technologies Of Aerial Multiview Photogrammetry

Oblique photography technology is a new and high technique developed in the field of surveying and mapping in recent years,by moving the camera tilt installation way to get more feature information,especially the facade of the building information.The aerial oblique image has become an important data source for obtaining urban regional information.It has made up the defects of incomplete image information and lack of facade information during the traditional aerial photography.It provides rich texture information for the fine modeling of buildings,and has a wide application prospect.The oblique photography has the following characteristics:a.It can acquire images with multiple perspectives and get more detailed ground information;b.It has high resolution and large field of view;c.Multiple images covering the same features have different resolution.d.The phenomenon of object occlusion is relatively obvious;e.The overlap of oblique aerial photography is greater than that of the traditional aerial photography,the more lens and larger overlaps lead to a mass of photos.This article studies the related technologies of oblique photogrammetry.The main tasks are as follows:1)Automatic extraction of multiple images’ connection points:For the characteristics of large inclination angles,high image overlap,and inconsistent image resolution for multiple images,SIFT feature extraction operators are used to automatically extract joint points.In addition,by using GPU parallel acceleration technology,the efficiency of image matching is greatly improved,and the SIFT operator’s computational complexity and memory consumption are greatly solved.The extraction of SIFT connection points has high precision and provides a reliable connection point for subsequent work.2)Aerial Triangulation:Due to the large tilt angle of the oblique image,the iteration calculation of traditional relative orientation does not converge,so using the multi-view geometry principle to solve the question.In addition,using vertical and oblique images for joint adjustment.3)Image intensive matching:Intensive matching is a key step in 3D reconstruction,using CMVS-PMVS method for intensive matching.This method is suitable for the 3D reconstruction of complex large scenes.It has good performance in reconstruction accuracy and integrity.The reliable and dense point clouds obtained provide a reliable basis for subsequent 3D reconstruction work.