Optical 3D measurement techniques for generation, revision and monitoring of digital twins of complex building structures

点击翻译按钮获取中文摘要

Optical 3D measurement techniques depict a powerful tool for the generation of geometry and system state information of digital twins of complex building structures. The potential of these methods, especially for digital twin data update and revision as well as for simultaneous determination of building monitoring parameters has only partially been exploited so far. In the project, optical 3D measurement methods for an efficient generation of precise digital twin 3D data will be further developed, employed and validated. 3D building representations obtained from image or laser scanner data will be enriched with building monitoring measurement data such as material strengths or crack patterns, obtained from the same image data by image analysis techniques. As a basis, conventional cameras or laser scanners will be used, which are either employed statically or in motion, including UAVs (unmanned aerial vehicles, ‚drones‘) as a sensor platform.On the basis of these image and laser scanner data, dense 3D point clouds will be generated by established photogrammetric techniques. In a next step, suitable geometric structures of a digital twin, given as CAD or BIM models, will be fit into these point clouds. This allows for the completion, improvement and revision of unprecise, incomplete or erroneous building representations. Moreover, additional building monitoring data, such as material strength or high resolution crack patterns, can be derived from the image data by image analysis techniques. Via the orientation parameters of the image data, these monitoring data can directly be transformed onto the digital twin, facilitating data interpretation by a user. Special focus will be put on the precise and reliable determination of material strengths from photogrammetric image and laser scanner data, requiring enhancements to photogrammetric data acquisition and processing schemes with particular attention to error propagation in large image blocks. As result of the project, photogrammetric methods will be available, which allow for an efficient and precise generation and revision of digital twins from image or laser scanner data, together with additional building state information.