🤖 AI Summary
To address the misalignment of cross-platform 3D models caused by the absence of geospatial metadata in ground- and airborne imagery, this paper proposes a fully metadata-free georegistration method. Our approach leverages only publicly available satellite imagery and a digital surface model (DSM), integrating neural radiance fields (NeRF) modeling with differentiable rendering to achieve robust geolocalization of airborne imagery via multi-view geometric optimization, and subsequently enables precise alignment of ground-level imagery to the airborne 3D model. We introduce, for the first time, a novel “satellite + DSM + NeRF” collaborative registration paradigm, supporting unified georegistration of non-overlapping and disconnected 3D models. Evaluated across multiple real-world scenes, our method achieves an average geolocation error of under 5 meters—without requiring any raw sensor metadata such as GPS or IMU readings.
📝 Abstract
Heterogeneous collections of ground and airborne imagery can readily be used to create high-quality 3D models and novel viewpoint renderings of the observed scene. Standard photogrammetry pipelines generate models in arbitrary coordinate systems, which is problematic for applications that require georegistered models. Even for applications that do not require georegistered models, georegistration is useful as a mechanism for aligning multiple disconnected models generated from non-overlapping data. The proposed method leverages satellite imagery, an associated digital surface model (DSM), and the novel view generation capabilities of modern 3D modeling techniques (e.g. neural radiance fields) to provide a robust method for georegistering airborne imagery, and a related technique for registering ground-based imagery to models created from airborne imagery. Experiments demonstrate successful georegistration of airborne and ground-based photogrammetric models across a variety of distinct sites. The proposed method does not require use of any metadata other than a satellite-based reference product and therefore has general applicability.