Automatic Co-registration of UAV-Based Photogrammetry and Terrestrial Laser Scanning in Urban Areas

Abstract

Laser scanning and Photogrammetry have become common techniques for providing monitoring services and producing 3D models for various applications in recent decades, including 3D modeling, reverse engineering, geoscience and anthropology, virtual reality, manufacturing engineering, and many more. A co-registration strategy is presented in this research to overcome the separated limitations of terrestrial laser scanners and Photogrammetry. The integration methodology used here is based on generating synthetic laser data to mitigate the extracting features problem for both 3D laser scanner and 2D Unmanned Aerial Vehicle (UAV) images. Then after, a structure-from-motion (SfM) procedure was applied following the automatic registration application. The fusion approach significantly reduced the roughness level of UAV images and resulted in more excellent density point clouds. This has a favorable impact on the level of detail obtained via fusion-based point clouds. Aside from increasing the finished model’s geometrical and graphical quality, this approach fixes the voids in Terrestrial Laser Scanning (TLS) data, retrieving more information, increases coverage, and assigning genuine colors to TLS data. This research represented a practical application in the urban city environment that seems typical towards future insights to make the compound a sample smart city based on geometry through a modified workflow starting from planning, data collection and processing, and data analysis and validation. Results have been analyzed statistically and discussed thoroughly for future co-registration development and application in multiple sectors.