I’m looking for workflow suggestions for the alignment of datasets in which the object position has been changed in relation to the background, without using masked images.
In this case I have 360 loops around a statue in a vertical position:
and another loop in the horizontal position:
I tried following the same workflow of the competition (without using masking), which consists of aligning the images separately in 2 components and merging them afterwards with control points.
I have set 6 points that are on at least 2 images in each component, and even with point weight set to 1000 I still get the usual Frankenstein model:
Has anyone managed to force RC to ignore the background in these cases or will only masking save me?
Ok, thanks for the feedback. I guess I can give up on these attempts.
I have these models already processed on the other software, I was just trying to see if I could use RC instead to do this (which I would prefer).
I also have the image masks on separate PNG files, but I checked if Photoshop could do some batch combination of JPGs and their alpha channels and it seems it doesn’t, which is unfortunate.