Closed ffurrer closed 7 years ago
This looks for the first at least two possible poses to do the alignment. A RANSAC approach that randomly selects two poses might be a smarter way though. At least if we have noisy data and not just outliers as in the current test.
Merging this without the RANSAC and will work on the RANSAC approach, once @mfehr has added the RANSAC for the time-alignment.
This looks for the first at least two possible poses to do the alignment. A RANSAC approach that randomly selects two poses might be a smarter way though. At least if we have noisy data and not just outliers as in the current test.