STL with low resolution?

[ascarneiro]

Hi,

Im using this software on trial mode, my need is export to STL and print on a 3D printer.

I scanned a small statue, and the final model on Zephyr is very good, take a look:

https://s22.postimg.org/p62i9d7z5/model3d.png

Then I go to EXPORT -> EXPORT MESH -> OPTION STL + DISABLE USE BINARY ENCONDING and the STL is saved. When I open the STL in any viewer or in printer software, the model apper with very low details, look:

https://s27.postimg.org/9nyh4tpib/stl1.png

I dont know if Im doing something wrong or if this is normal. My goal is generate a more detailed STL to print.

So, it's possible do something better?

Thank

Alvaro

[Andrea Alessi]

Hi ascarneiro!

the STL format does not contain triangle connectivity information (which usually other formats do). This means that often, normals might be slightly different. This is why the lighting effect of the STL mesh you export might differ a bit from what you see inside zephyr.

Strictly speaking of geometry however, you should see no loss of detail when exporting.

Regardless, if you'd like to share with us the zep file, i'll gladly have a look at it!

Andrea

[ascarneiro]

Ciao Andrea!

Grazie per avermi aiutato!

I uploaded the file compreesed here:

http://acplanet.net/projeto.zip

Let me know your finds.

Thank

Alvaro

[Andrea Alessi]

Ciao Alvaro, nessun problema

Yes, it's simply shading. You can double check yourself by exporting your mesh for example in .ply, open it in meshlab and compare the output by disabling color and switching from flat shading to smooth shading. It's just a matter of how a software shows a mesh with certain lights.

HOWEVER =)

It looks like you simply cut the knight mesh from the bigger mesh. In fact, you only have a dense point of all the room in your workspace! In some cases zephyr can recognize also the room, but since zephyr can't know what you want to reconstruct, it computed a bigger bounding box. This means that the discretizaton is lower, and in fact, if you see your dense point cloud you'll see that there aren't many point on the knight, right?

We can fix that! f this happens, simply rescale the bounding box after you compute the sparse point cloud, so that the generated dense point cloud is of the subject only. I can guarantee you you will have a much much better discretization there!

So in your case i would compute another dense point cloud (i see that you rescaled the bounding box,but that was done after the mesh generation) using a smaller bounding box. Then generate a mesh of the knight. That should help a lot.

Another solution would be to mask the images, but that can take a bit of time.

I hope i explained myself, as it's friday

If you're still confused though i'll gladly process it for you and show you how you can do what i meant. But i also would need the photos if you can share them with me as well! (on a side note, the acquisition was perfect!)