3D Modeling And Accuracy Evaluation Of UAV Inclination Photogrammetry Based On Smart3D Software

As the UAV photogrammetry technology becomes more and more mature,the operation mode of constructing a three-dimensional model based on oblique photogrammetry is playing an important role in smart cities and other fields.Oblique photogrammetry uses multi-camera systems to capture images from multiple angles of the ground and acquire precise texture information on the side of the ground objects.The use of photogrammetry software to process acquired multi-angle images can ultimately generate realistic three-dimensional models with real-world texture information,which have been widely used in various fields of the national economy.However,the problem is that the construction of three-dimensional models based on oblique photogrammetry technology cannot accurately evaluate the accuracy of the three-dimensional model.The reason is that the three-dimensional model constructed by oblique photogrammetry and the traditional three-dimensional are different.The traditional three-dimensional model adopts an artificial construction mode,and is modeled layer by layer under the assistance of topographic maps and other data.The accuracy of the three-dimensional model can be accurately measured by measuring its internal parameters.However,due to the particularity of the data form,the modeling method based on oblique photogrammetry can not select the model data and cannot accurately assess its internal accuracy.Therefore,the accuracy of the three-dimensional model was evaluated by measuring the coordinates of the control points and the measured values of the model in the field.In this paper,the remote sensing production validation and extension research project of the domestic satellite data of Sichuan Provincial Department of Land and Resources in the land use monitoring and status change is a technical support,and a part of the main campus of Chengdu University of Technology is taken as the research area,based on the oblique camera measurement of drones.The technology acquires multi-view images,followed by in-depth analysis of the key technologies and algorithms for oblique photogrammetry modeling of UAVs,and then studies and explains in detail the deployment and acquisition methods of image control points,and finally uses the oblique photogrammetry processing software Smart3 D to solve the problem.The study area multi-view images,and finally built a three-dimensional model.At the same time,by measuring the coordinate data of the control points and the relevant building parameter information in the field,the accuracy of the actual measured values of the model was analyzed and certain results were achieved.The main research contents and achievements of this article are as follows:(1)A systematic and in-depth analysis of the current key technology for oblique photogrammetry,including image matching,aerial triangulation,multi-view image dense matching,TIN network construction,texture mapping,etc.,and selected Smart3 D as the oblique photogrammetry modeling software.(2)Analyzed the current algorithms based on UAV oblique photogrammetry modeling software,including SIFT feature extraction and matching algorithm,structure from motion algorithm,multi-view clustering algorithm,patch-based dense matching algorithm.(3)Main research results: Using Chengdu University of Technology as a research area in some areas of the main campus,a high-resolution multi-view image of the study area was successfully obtained by using a light UAV oblique photogrammetry system;Obtained the real three-dimensional model by using UAV photogrammtry technology;Actually measured the control points data,related building parameters,etc.in the field,and calculated various errors of the three-dimensional model through the relevant formulas.The result proves that the three-dimensional model generated in this paper meets relevant specification standards.