At the company I am working, we are recreating surgical scenes with photogrammetry and laser scans. Previously we created the photogrammetry and laser scan separately and then merged those. But i would like to try merging those two from the beginning.
We have a Creaform Go!Scan 3D Spark laser scanner, and it can do the following formats for export:
obj, stl, 3D points text file txt, .zpr, .wrl, .ply, .x3dz, .x3d, .ma, .dae, .fbx, .3mf. .txt, 3D points normal text file .txt, iges,
Any suggestions how to solve this?
Why I need this: we need as accurate geometry as possible, that’s why we are using a laser scanner, but the fatty, shiny parts are invisible for the scanner, so we would like to fill the holes with photogrammetry
Correct me if I am wrong, but that laser scanner is a mobile scanner which creates unordered laser scans? Were you able to use laser scans from that scanner before? If yes, how did you make ordered scans?
We recommend using one of the methods mentioned in our tutorial:
And one of those methods is to create separate components and merge them. Merging data sets immediately without creating components separately requires a really good overlap between images and scans (that is the main requirement).
Now I can see, that my phrasing was not correct. Your first questions are the ones, I would like to know the answer for also :)
I am not sure that I am familiar with the terms, ordered/unordered. But yes, it is a mobile handheld scanner, creating live reconstruction. Meaning, that the mesh is created continuously while you are using the scanner above the object. I think that’s the reason I cannot export a point cloud format, like for example from a BLC scanner for revit.
We used scans from this scanner before, but not with reality capture! We were exporting .obj, and we used meshroom for photogrammetry. After that, we used geomagic, blender or meshlab, to merge those.
We tried reality capture now, and we can see that it is way more advanced then meshroom for photogrammetry. I am just curious, if I can merge that with our laser scans, inside reality capture, so we can make the process easier.
So basically, I am stuck at importing the files I have into reality capture because of the file formats. That’s the main problem.
I tried the .ply (ascii and binary too) but it says “the property type “list” is not supported”. The other ones are not even on the list for import.
For laser scan you can use PTX of E57 formats, but scans have to be ordered. This means that the positions of the laser scanner have to be known - this is the reason why mobile laser scanners are not supported. This is something that is being worked on, and hopefully mobile scanners will be supported in the near future.
This article will maybe give you more information:
Laser scans when imported are being converted into a cube of 6 images (cameras) to mimic the photogrammetry principles. And because of that, the scanners position is required, otherwise RealityCapture won’t know where to put the ‘cameras’ and scan will be empty.