Accuracy Optimization and Engineering Applications of 3D Modeling for Complex Terrains based on UAV Oblique Photogrammetry

Abstract

With the occurrence of UAV oblique photogrammetry, the collection of geographical information has been greatly optimized especially when it comes to creating 3D models of complicated terrains. It's got clear advantages over ordinary survey techniques like terrestrial laser scanning, manned aerial photogrammetry when it comes to being cheap, easy to operate, and providing detail. But getting the 3 models created of intricate landscapes with steep slopes, lush plantings, great rises and falls, and city canyons all to be very accurate is still very tough. This paper gives a thorough review along with detailed analysis of the essential technologies and methods for improving the exactness of 3D models which are obtained with UAV oblique aerial photographing method in complicated terrain. It covers the whole work flow all the way up to including the best flight missions with the most strategic data and using the latest data processing tools. Main areas include sensor characteristics and tough camera calibrations, the clever setup and location of Ground Control Points (GCPs), and how mixing direct georeferencing system information such as Real-Time Kinematic (RTK) and Post-Processed Kinematic (PPK) positioning data really transforms things. In addition, the paper looks at the meaningful influence of multiple source data synthesizing, combining oblique images with information obtained from sensors like Lidar, terrestrial photogrammetry etc to solve occlusion and to improve the completeness and verisimilitude of the model. analyze all sorts of factors affecting model's accuracy and present a number of mixed improvement approaches. The practical utility and effectiveness of these optimized models are illustrated via an in-depth review of their engineering applications which incorporate geological disaster monitoring, open-pit mine surveying, civil infrastructure inspection, as well as natural resources management. Four detailed tables, and quantified the improvement accuracy, and demonstrated the practical value in real-world project engineering, which could transform such technology into real-world project engineering.