Automated UAV image-to-BIM registration for planar and curved building façades using structure-from-motion and 3D surface unwrapping
Cheng Zhang, Yang Zou, Feng Wang, Johannes Dimyadi
Abstract
Texturing Building Information Model (BIM) with up-to-date Unmanned Aerial Vehicle (UAV) images has brought substantial benefits to building façade inspection. However, current image-to-BIM registration methods are sensitive to UAV positioning accuracy and façade features. Additionally, perspective and geometry distortions on UAV images hinder the texturing of curved façades. To address these issues, this paper introduces an approach to registering UAV images onto BIM for buildings with both planar and curved façades. Firstly, Structure-from-Motion is used to align UAV images with BIM. Secondly, fragmented and distorted UAV images of planar and curved façades are converted onto a distortion-free panoramic image using 3D surface unwrapping. Thirdly, the generated panoramic image is projected onto BIM surfaces as textures. The process has been automated using a Dynamo prototype and evaluated through computer simulation and a real-world case study. The outcome shows a mean accuracy of 1.7 pixels and 8.5 mm.