Looks like most of the 3D modeling packages do not directly accept a point cloud data set from a 3D scanner like device. What could be the best strategy for converting a point cloud data set, especially one which is large, into a 3D model?
Thank u.
I could locate some specialized application (3D Reshape) based on your suggestion
Have a look at the free version of GOM inspect. It should do the requested task.
http://www.gom.com/3d-software/gom-inspect/features.html
As in many aspects of life the best strategy depends on the desired results and on the assets that you possess, I am sure you know that. :) From my experience the most important parameters that count are:
- number of point clouds that represent the object (one or more),
- merging/registeration quality of point clouds if more than one,
- surface quality in a single point cloud (noise and other errors),
- can the process be human-supported or should it be completely automatic,
- is the model to be used for visualization, inspection or other applications,
- does the model consist mainly of shape primitives or is it a free form?,
- reqiured accuracy of representation.
In our applications (3D measurement for cultural heritage archiving) we mainly register hi-res point clouds based on ICP and create triangle meshes for visualization. If density of point clouds is sufficient we often use these for direct visualization (people working at museums are often hypersensitive to any additional data processing).
If you have or can developp a mathematical model for your object, with parameters, one strategy may be to use the point cloud to fit model parameters , by minimizing the rmse.
You can also take a look at the PCL library (point cloud library) and also at the MeshLab software application. They can both do what you want using different methodologies.
Check out our SSD surface reconstruction open source software
http://mesh.brown.edu/ssd
- Badges
- Science topic
- Similar topics
- Correction