Stochastic Reasoning for Uav Supported Reconstruction of 3d Building Models

Abstract

ABSTRACT:The acquisition of detailed information for buildings and their components becomes more and more important. However, an automatic reconstruction needs high-resolution measurements. Such features can be derived from images or 3D laserscans that are e.g. taken by unmanned aerial vehicles (UAV). Since this data is not always available or not measurable at the first for example due to occlusions we developed a reasoning approach that is based on sparse observations. It benefits from an extensive prior knowledge of probability density distributions and functional dependencies and allows for the incorporation of further structural characteristics such as symmetries. Bayesian networks are used to determine posterior beliefs. Stochastic reasoning is complex since the problem is characterized by a mixture of discrete and continuous parameters that are in turn correlated by nonlinear constraints. To cope with this kind of complexity, the implemented reasoner combines statistical methods with constraint propagation. It generates a limited number of hypotheses in a model-based top-down approach. It predicts substructures in building facades -such as windows -that can be used for specific UAV navigations for further measurements.

Identification and Modelling of Translational and Axial Symmetries and their Hierarchical Structures in Building Footprints by Formal Grammars

“…Our method has been successfully applied for generating façade hypotheses for CityGML LoD3 buidlings (Loch‐Dehbi et al ). Our method delivers good prior knowledge accelerating the reasoning process yielding a small number of accurate hypotheses and hence reducing the search space significantly.…”

“…In order to achieve a good quality interpretation, models providing prior knowledge are needed. While most approaches rely on high resolution input data, such as dense 3D point clouds or images, novel methods try to generate semantic models based on sparse observations like footprints and strong priors like probability densities and constraints (Loch‐Dehbi et al ). Footprints are widely available, at least in developed countries, from cadastrial data or from Open Street Map.…”

“…The inferred rules can be used as background knowledge in order to predict building parts such as occluded parts which are unknown beforehand. Furthermore the grammar rules enable a compact representation of buildings and their parts especially of symmetric ones.The fact that symmetric building parts are often constructed in a similar way helps to restrict the space of hypotheses during a stochastic reasoning process for reconstruction of 3D building models as described in Loch‐Dehbi et al (). Besides, our approach detects symmetries and infers adequate grammar rules even though symmetry is broken in the given footprint (cf.…”