Intelligent Application of Photogrammetry in Computer Reverse Modeling of Building Structural Engineering

Abstract

This paper addresses the challenge of inadequate elevation information acquired through oblique photogrammetry in building modeling. The study explores the application of oblique photogrammetry for reverse modeling, focusing on close-range photogrammetry in the context of building structural engineering. The methodology involved the use of a consumer quadrotor UAV, specifically the DJI Phantom 4 Pro, equipped with a tilt camera on the flight control platform. The UAV captured image data, and the experimental setup included checkpoints on the keel frame to assess the accuracy of the 3D model. The experimental findings reveal that the disparities between measured plane distances and model distances range from a maximum of 0.006m to a minimum of 0.003m. Additionally, the maximum difference in height measurement compared to the model is 0.005m, while the minimum difference is 0.002m. Maintaining accuracy at the millimeter level showcases high geometric precision. The unmanned aerial vehicle tilt photogrammetry method works as expected when used in real life, meeting the needs for individual modeling and providing more detailed model facade textures with higher resolution. The study validated the effectiveness of the unmanned aerial vehicle tilt photogrammetry method by achieving accuracy at the millimeter level and high geometric precision. The study showcased its ability to meet relevant specifications for individual modeling, providing detailed model facade textures and enhanced resolution.