Creating 3-D Maps of Complex Buildings for Disaster Management

Creating 3-D Maps of Complex Buildings for Disaster Management

Disaster managementCreating 3-D Maps of Complex Buildings for Disaster Management


Published 23 October 2020

In case of an emergency, first responders like the fire brigade need up-to-date information. 2D maps are a common source of information but they can be difficult to read in an emergency situation. A researcher has created an algorithm that can accurately generate 3D models of the insides of large buildings from point clouds.



In case of an emergency, first responders like the fire brigade need up-to-date information. 2D maps are a common source of information but they can be difficult to read in an emergency situation. University of Twente’sPhD-student Shayan Nikoohemat created an algorithm that can accurately generate 3D models of the insides of large buildings from point clouds.


Indoor 3D models are the digital twins of building interiors. The 3D models could be used by first responders to get a good impression of large buildings, like shopping malls, a hospital or a sports complex, on their way to the emergency. 2D maps represent important information – like the location of emergency exits – on tangled floor plans, making them difficult too read quickly and after each reconstruction, these maps are outdated. “Sometimes, these maps are so outdated that the real building looks completely different than the floor plans. We need a fast and reliable approach to create the digital 3D model of interiors.”, says Shayan.


From Point Cloud to 3D Map
Luckily, laser scanners can quickly scan a whole building after every reconstruction. However, these scanners create point clouds, unstructured data which still has to be converted into a 3D model. The data doesn’t know if a scanned point is a wall, an exit or, for example, a table. According to Shayan, his program solved this: “For my PhD thesis, I created algorithms that automatically understand the data and can create 2D and 3D maps. We can detect and model doors, stairs, obstacles and navigable areas which are crucial data for the emergency planning.”


Recognizing Elements
The algorithm can recognize different structural elements such as walls, slabs, ceilings, and openings. Individual items like furniture, however, still pose a problem. “It is not yet able to correctly label everything, but the structural elements are enough to create an accurate map, which we tested on several real datasets,” he says. During his postdoc, he will further develop the system to also work for individual items. Huib Fransen of the Safety Region Rotterdam-Rijnmond was delighted with the results: “Shayan’s project is exciting for us and we were happy to provide him with the scanning sites for test cases.” 

Leave a comment


Register for your own account so you may participate in comment discussion. Please read the Comment Guidelines before posting. By leaving a comment, you agree to abide by our Comment Guidelines, our Privacy Policy, and Terms of Use. Please stay on topic, be civil, and be brief. Names are displayed with all comments. Learn more about Joining our Web Community.