I am seeking help to mesh complex point clouds create from topogrophic images of surfaces and coatings.
Some context:
I am a material physicist working in acadamia and I often come across 3D images of surfaces used for roughness and topography analysis (some example would be atomic force microscopy images, 3D profilometry or surface simulations). For educational purpose I tried to 3D print these type of data to illustrate phenomenon and physics principles to students.
The problem:
The analysis software I have access to (open source Gwyddion and proprietary Vision64 from Bruker) do not allow saving in standard formats for 3D printing (stl, steps etc).
These software allow me to save in a .xyz format which can be used to generate a point cloud in Meshlab. However I am not able to go from this point cloud to a fully meshed model.
Here is my workflow (based on several blog posts on the meshing of landscape data):
1- Generate the point cloud from the analysis software (.xyz file) (original image Figure A)
2- Import in Meshlab (see Figure B)
3- Compute normal from point set
4- Meshing, this is where I am stuck:
For the meshing I tried to use the poisson reconstruction but it doesn't seem to be able to accomodate the fine and highly rough surface (see Figure C).
I have also tried the voronoi reconstruction and this seems to capture more adequatly the expected surface but with a lot of holes in the mesh (see Figure D).
Any thoughs and advices on how to proceed to get a correct mesh?
Disclaimer: I am quite new to the world of meshes, 3D models and digitalization.
Last edit: Gerixeti 2018-09-23
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I am seeking help to mesh complex point clouds create from topogrophic images of surfaces and coatings.
Some context:
I am a material physicist working in acadamia and I often come across 3D images of surfaces used for roughness and topography analysis (some example would be atomic force microscopy images, 3D profilometry or surface simulations). For educational purpose I tried to 3D print these type of data to illustrate phenomenon and physics principles to students.
The problem:
The analysis software I have access to (open source Gwyddion and proprietary Vision64 from Bruker) do not allow saving in standard formats for 3D printing (stl, steps etc).
These software allow me to save in a .xyz format which can be used to generate a point cloud in Meshlab. However I am not able to go from this point cloud to a fully meshed model.
Here is my workflow (based on several blog posts on the meshing of landscape data):
1- Generate the point cloud from the analysis software (.xyz file) (original image Figure A)
2- Import in Meshlab (see Figure B)
3- Compute normal from point set
4- Meshing, this is where I am stuck:
For the meshing I tried to use the poisson reconstruction but it doesn't seem to be able to accomodate the fine and highly rough surface (see Figure C).
I have also tried the voronoi reconstruction and this seems to capture more adequatly the expected surface but with a lot of holes in the mesh (see Figure D).
Any thoughs and advices on how to proceed to get a correct mesh?
Disclaimer: I am quite new to the world of meshes, 3D models and digitalization.
Last edit: Gerixeti 2018-09-23