I have a fairly big wodden model which been split into 3 parts in order to document it easier. Please look at enclosed picture. I´m now at the state finding a solution to merge all 3 into 1 model in RC. I guess that using controll points won´t help in alignment. I´ve exported all 3 models as RC ALIGNMENT COMPONENT and then loaded again in a new project. So far so good but how to align my 3 models in a row into a new one. The real big physical model have maybe a few mm in between sections but isn´t a issue if the 3D model is almost tight together.
Forgot to mentioned that I can always go a detour in other software, like Maya and import all 3 models but a very long time need to be spent in UV map the whole model again. Not an option for me
I think the only chance will be to define Ground Control points with known coordinates and then orient all components with the option “Merge Georeferenced Components” set to true. You could use the corners of the model, where the parts join up and measure those with a tape and then create the coordinates from those measurements…
It makes sense if you get familiar with the basics on your own (which it seems like you did). But if you are stuck with a specific problem, this is what the forum is for. And afaik this is the first time somebody asked for a solution for this kind of problem, so it’s interesting for everybody I guess…
I still struggle when merging 3 components into one model. I´ve defined the CP:s in meny as Ground Control (not tie points) and new values in Coordinate system. After running alignement (georeferenced), the 2 halves become one almost perfectly but still one side is still a little bit rotated off center. I really want point 10 and 11 flush together. Tried to set distance between those point to 3mm but nothing happens. It´s like when you design in 3D CAD, take one corner and merge with another face, corner or edge. This would be helpful to do in RC
Anyone got an idea
Are you certain that 10 and 11 have the same coordinates?
If yes, then I imagine that RC is still trying to force the model into a shape it thinks is right, meaning that the background might have more and better tie points than the model itself.
In any case, texturing completely separate parts in one model will only result in chaos.
So I guess what I wouls do is to reconstruct and texture all of the models separately and then assemble them with external software…
Well, I think this case is worth making a tutorial of it … if I ever make the point …
I’ve got around 10 georeferenced models within one UAV scan with ground taken pictures ( it creates around 100 components at alignement and I only keep the geo-taged ones )…
GPS coordinates differs a little between those models so I assume if I get precise coordinates at building angles from Model#1 ( UAV, the wider ) and I report those coordinates ( euclidian ) at control points for other models at the edges of building or noticeable things It will all blend okay ?
What about the non- geo tagged components ? might I apply here the same solution ?
