Optical 3D measurement methods for generation, updating and measurement data acquisition for digital twins of complex building structures

The project deals with the conception, implementation and validation of photogrammetric methods for the efficient generation of highly accurate 3D geometric data for digital twins of complex building structures as well as the enrichment of these representations with building monitoring parameters. Measurement and inspection data such as material thicknesses or crack patterns are derived using image analysis techniques applied to data from cameras (Figure 1) or laser scanners, which are used statically or in motion, both terrestrially or on UAVs (unmanned aerial vehicles).

From the image or laser scanner data, 3D point clouds are generated (Figure 2) using established photogrammetric methods, wherein suitable geometric representations of a digital twin are then fitted.

Based on this data, inaccurate, incomplete out-dated or erroneous as-built data can be complemented, improved and updated. As a by-product, the orientation parameters of the captured images are determined, also facilitating visual interpretation. Automated image analysis procedures are used to generate high-resolution crack patterns from image data, which can be transformed directly into the geometry of the digital twin using the orientation parameters of the captured images. Condition parameters such as surface roughness or camber can also be derived from the data. Special attention is given to the determination of material thicknesses (exemplarily shown in Figure 3, right) from image and laser scanner data, which requires specific modifications of photogrammetric data acquisition processing schemes with a special focus on error propagation in large image data sets.