Digital measurement: Berlin starts revolutionary road construction project!
Digital measurement: Berlin starts revolutionary road construction project!
Berlin, Deutschland - Berlin plans to recording new geodata in the coming months. For this purpose, special vehicles are used, which are equipped with state -of -the -art cameras and laser scanners. These vehicles will absorb 360 ° panoramic images and 3D point clouds of the street space. The aim of this measure is to accelerate infrastructure planning, since the existing geodata are largely out of date and new recordings are urgently needed. State Secretary Johannes Wieczorek (CDU) emphasizes the importance of this digital inventory for future infrastructure planning and emphasizes that individual measurements should be omitted in order to save both time and costs. Access to the data obtained is provided via a geoportal. In order to maintain privacy, recorded vehicles or faces are anonymized. This reports rbb24.
In parallel to these developments, the Federal Office of Cartography and Geodesy (BKG) has advertised a research project that deals with the automated processing of 3D point clouds using artificial intelligence (AI). The aim of this project is the development of methods to derive objects large and robust from comprehensive laser scan beams. It is to be examined which additional information from 3D point clouds can be obtained and how exactly automation is possible. This research has special relevance in the context of digital twin Germany, a concept that includes digital representations of real systems. The BKG is looking for research institutions or companies that have expertise in the AI-based analysis of 3D point clouds and are able to integrate them into digital infrastructures. BKG explains the requirements for application.technologies and methods of data processing
For the efficient measurement of large structures, as often occur in urban space, there are considerable amounts of data. Methods of artificial intelligence are used here, which are used to optimally prepare data from laser scanners or cameras. This data is crucial for applications such as Building Information Modeling (BIM) and Augmented Reality (AR). Point clouds have established themselves as one of the most common data formats in 3D measurement, but require extensive preliminary processing. In order to interpret these large amounts of data sensibly, the researchers use various strategies for data reduction and interpretation. A semantic understanding of the point clouds and image data is required to recognize relevant objects and locate correctly. According to Fraunhofer IPM
In order to implement these technologies, powerful hardware is necessary because the training of such neuronal networks requires a high storage requirement. High -performance servers are available to enable the parallelization of arithmetic processes and to make the processing of large amounts of data efficiently. The advances in AI-based data processing could not only increase efficiency in infrastructure planning, but also significantly improve the quality of the geodata recorded.
| Details | |
|---|---|
| Ort | Berlin, Deutschland |
| Quellen | |