Problems working with stl/obj files from scan data

[system]

Good morning,

Everything I do starts with CT data which is segmented and then modeled and exported as STL files. I clean them, decimate them, and ensure they are a single shell and then import into ZBrush 4R6.

The first issue is the absolutely miserable effort I have to go through to get the files lined up so I can use symmetry properly. I am using a lot of skulls, so symmetry is never going to be quite right, but it should be close. After the use of two or three other programs, including MeshLab, it comes into ZBrush. I need to ensure the geometric accuracy of the resultant part, so the imported model appears huge in ZB. Most skulls with mandible are in the neighborhood of 200-240mm vertically.

In attempting to follow the technique of using Polypainting on the surface, separating the marks, then extruding them to create tissue depth markers, I find that nothing I do actually paints on the surface of the model. I have been using the technique by Ryan Kingslein.

Is part of the issue the fact that the model is a highly tessalated, triangle based model rather that quads? I am only going to place muscles and skin on the model for facial reconstruction, but since I have many to do, it would be nice if I didn’t have to find a way to reconstruct the skull too.

Any suggestions?

[Doug_Jones]

ZRemesher can get you mostly quads with a couple clicks. Not sure why you can’t paint though. What is final output from ZBrush for, still image, animation in a game?

[marcus_civis]

The issue of triangles should only be an issue for sculpting or if the model is not very dense, polypaint (vertex color) might not be of good definition.

The most likely cause of your problems is the size. Many of ZBrush’s functions work best if the model is quite small (about 2 generic units). ZBrush will auto-scale OBJs imported into the Polymesh3D Star but the STL import maintains the original size.

You could try reducing the size of the model after import. One method is to use the Tool>Geometry>Size XYZ slider. If you note down the original value you can restore it before export.

[MentalFrog]

Export as OBJ instead of STL if you can. That should help maintain the scaling issue. As for the painting like was mentioned use Zremesher to put the model into a low quad only model and then sub-divide as needed.

You could try reducing the size of the model after import. One method is to use the Tool>Geometry>Size XYZ slider. If you note down the original value you can restore it before export.

A much better way is to change the Tool>Export>Scale number. It’s basically the same thing Marcus is saying but you don’t have to write anything down and Zbrush should automatically export back to the orginal size based on that number. If your models are about 200-250mm then a good number would be about 100 as the export scale. I do a lot of models in the 75-175mm range and I usually use 50.

I work with scan data all the time so if you have any questions let me know.

[system]

I hate to sound like a moron! The software used to segment the data from CT scans is Mimics by Materialise. I will try exporting in ply or iges next time. Since the normal use of the files is to create a physical model, the stl is the default output file type.

I have been printing the output for about 12 years, but the job I am in currently needs to do more sculpting and reconstruction on the resultant file. Not normally sculpting directly from the file, but ON the file, ie. facial reconstruction on the skull. To do this in an acceptable forensic fashion, this requires the placement of tissue depth markers in many places, both down the midline and bilateral symmetrically. Ryan Kingslein has a very nice tutorial on Anatomy of the Face. After placing the markers, each of the muscles is laid down to the correct thickness and then skinned. Most people are using self-modeled skulls. I need to use actual reproductions, which will then be printed.

Big problem! Scanned skulls are completely accurate - complete with holes in all the right places (but not holes in the ‘mesh’ sense), and at times the bone is thin enough or damaged, so holes in the wrong places are there. I have no clue how to get these holes to be closed surfaces so I can derive just an outer skin model of the skull - with no openings or place for the brain!

Since this is for medical forensic or museum use, I need to be quite careful to end up with a finished product that is sized correctly.

Can you clarify for me once more how the scale function works? I am attaching a jpg of the one I am currently working on. I will try to follow the suggestions and see how it goes. Thanks for all your help!

Attachments

[system]

When doing as suggested, and reducing size to 25%, I now get completely separated. Should I merge the two pieces before size reduction? I am attaching 2 jpg’s. One is at 25%, the other is merged, then reduced, and a small area painted. More like polygons painted! What gives now?