Could we prioritise, say, the latest scan, such that if it overlaps an area covered by another, older scan, that its data is used in preference, and the 'merging' only happens at the edge of the new scan, where it blends into the old scan?
I have several scans of a dive site, and some of the scans overlap quite a lot, and have been done over weeks/months... plus, some of the terrain is sand, which moves from day to day, the ripples are different, or it'll be 0.5m different height after a few months or strong swells.
So when i merge the scans from each, almost all of it is great, as the rocks an corals haven't change at all, or not enough to care about... But at the edge of the reef where it meets the sand, i have problems with two surfaces, one floating above the other, showing the old level of the sand and the new level.
Right now, i'm facing the task of going thru each of my scans and disabling all the cameras in the older scans that cover these sandy areas, problem areas, etc, such that when i merge all the scans photos together for the big Alignment, i won't have too many weird surfaces to deal with. This may take i while i think.......
IF we could give, say, Camera Groups hierarchy, so each mission would be in its own group, and you'd number them as needed, possibly chronologically, or just where one mission got better date...
Then when aligning, MS would take that group's cameras to form the topography, ignoring most of the overlapping group with the lower ranking.
Obviously, each scan track goes a bit fluffy at its edges, so perhaps it'd have the boundary set at a threshold value, so that it was still strong data before it blended into the old scan data.
As i'm doing my mapping/modelling project, i'm learning a lot about the techniques needed in the scanning, and i'm also sometime presented with much better conditions, visibility and lighting, such that the scans gotten are just much higher quality.
This would make adding higher quality scans to a data set much, much easier and quicker, for us multi-mission folk.
Topic: multi-mission scans- give priority to latest, when overlapping? (Read 1204 times)
