Is there a set of functions (workflow) to compare (superposition) an original model to a scanned 3d mesh? This is one of the most important and basic quality evaluations, but I haven’t seen a good lesson on how to accomplish this.
Best
Is there a set of functions (workflow) to compare (superposition) an original model to a scanned 3d mesh? This is one of the most important and basic quality evaluations, but I haven’t seen a good lesson on how to accomplish this.
Best
Hi John, Do you have a small example file you can share of the conditions?
the file is a client file, so no, but the ideally the problem is to compare the distance between 2 aligned meshes or a mesh and a drawing. I could create a small 3d object on my Bambu and scan it if it will help.
If it was my task, I would align two volumes by their centroids, then I’d populate scanned mesh with points or used vertices, then I’d measure the distance from point to the original geometry (Pull points) looking for the greatest value or average, or whatever you are interested in. Distances can be evaluated with colour for better visibility.
Fairly easy for Grasshopper.
Cloud Compare (free) and GOM Inspect can do this. It might be possible to find a free version of GOM Inspect around somewhere. Zeiss has picked up GOM inspect and I don’t know if they offer a free version of the latest release.
Hi John -
Have you tried PointDeviation?
-wim
Great responses. No, I have not tried these tools. Cloud compare and the suggested grasshopper flow look promising. Id like to do it in grasshopper as it has the possibility of extending into acceptance criteria. The query model function gets the points and the mesh closest point make intuitive sense.
I think the alignment is the challenge as there is no shared or register coordinate system. When I import the scanned mesh, is there a function in Rhyno to “move to” 2 or 3 selected points defined in both objects (e.g. a corner). That could slide it into rough orientation. There may be some form of center mass function as well…
You could use a bounding box or a volume centroid to help align things.
Cloud Compare has a fine registration tool that does not require picked points. It produces a best fit based on a sampling of all the points. It does not use the centroid but rather the proximity of the points. GOM Inspect has a similar best fit function, and can also allow constraints to be defined, but the interface is poor for the constraint application.