Details Regarding Ground Truth

Extensive and time consuming efforts were made to groundtruth the logs available here. In summary we aimed to provide groundtruth for each pose across the entire building using as built floor plans to within 2-3cm (for each pose).

The ground truth proceedure is as follows. For a small batch of laser poses (e.g. 160 scans or 4 seconds), we align the start and end scans to the floor plan and carry out incremental LIDAR scanmatching in between. We then construct a small pose graph optimisation problem with the Factor Graph illustrated below.

[MF: put figure here]

Relaxing this posegraph (using iSAM, developed by Michael Kaess) produces the final groundtruth poses for that small batch. This process is repeated for each subsequent batch of scans. The scanmatching mentioned above uses the Fast and Robust Scan Matching (developed by Abe Bachrach) which produces very low drift rates in all situations (except long corridors).