Heterogeneous sensor data fusion for urban digital twin modeling
No field work required, any programming skills
Normally, building digital twins are based on data collected beforehand from one sensor mounted on an aerial or ground platform. ALS can offer a large amount of data suitable for mapping, however, data is often incomplete since there is mostly no information about ground level, and often around the facade or structures which have been covered by other occluding structures. From the ground, mobile mapping systems MMS can also offer city-wide coverage with a high level of positional accuracy but have the limitation of only mapping urban features along the roads. Accordingly, fusing both heterogeneous sensor data from the air and the ground has the advantage of having a complete coverage model of reality. This is still a challenge in terms of data dissimilarities and the need for co-registration.
In this research, two different data are taken from the ground and aerial sensors, namely: Airborne Laser Scanning ALS and Mobile Laser Scanning MLS will be coregistered to lead finally building an urban city digital twin.
The coregistration of both sensor data is expected either to be applied geometrically or optionally implies deep learning techniques for the corresponding urban features segmentation.