These sensors were equipped on a tablet to allow portability and real-time visualisation of the results produced. The data generated by the sensors was then processed using SLAM (Simultaneous Localization and Mapping) technologies to obtain a final result consisting of a 'coloured point cloud' like the one in the image. The sensors involved are inertial sensors (IMU), traditional cameras (mainly fish-eye lens) and 3D ToF (Time of Flight) cameras for the final mapping. As the cameras have a limited FOV, it was necessary to move the tablet in the environment to obtain a complete and accurate reconstruction. At this stage, it was crucial to correctly estimate the path of the tablet within the space by means of the installed sensors. To validate the technology, the results produced were verified in the laboratory by comparing them with those produced by a MoCap system available at the Spatial Computing Lab of the Institute of Information Systems and Networking (ISIN).

The Master's student participating in this project worked on the development phase of the SLAM methodology and the optimisation of heuristics to manage the process in real-time. They also helped to develop the validation part of the results using MoCap, which required the synchronisation of devices (tablet/Mocap) using NTP (Network Time Protocol) type protocols. Finally, they participated in the automatic deployment of the solution produced, which was put into production on dozens of the project partner's devices.

 

Conclusions

Excellent results were obtained, with a maximum error in estimating the final trajectory of 5cm under the laboratory conditions tested and final reconstructions validated both metrically and visually. The student was able to put what was produced to the test under various application conditions; it is worth mentioning the complete mapping of the Villa/Museo Vittoriale degli Italiani, the house where Italian poet Gabriele D'Annunzio lived his final years. The photo above depicts the bathroom of the villa.