I’ve tried using Feature Edges and Delaunay2D, but they don’t give me the result I’m looking for. Does anyone know how to achieve an outline that matches the exact perimeter of the irregular data points?
Apologies, I’m still a beginner with ParaView and VTK visualization. Does this mean that my VTK file should be in ASCII format instead of binary? If possible, could you share a resource or guide on how to properly convert it to the correct format?
Additionally, could you provide a VTK file with points that I can use with the Feature Edges filter? I’d like to share it with the developer of the Tomofast-x software so they can consider implementing it in a future update.
The challenge of your data is that it what we sometimes call a “point cloud” or “mesh-less data”. Your data does not define a mesh of cells connecting points and defining the area of space it covers. That means that ParaView does not immediately know the area that is supposed to be covered.
There are some tricks to impose a mesh on a point cloud. I think the one that will work best in your case is to use the Delaunay 2D filter. This filter will connect together the points in your data with triangles.
One problem you will see at first with the Delaunay 2D filter is that it will form a convex hull of the points, which means those cut out areas will get filled in. The Delaunay 2D filter has an optional parameter named Alpha that is used to remove stretched out triangles that tend to cover those spaces we want to see empty. You have to play with that parameter a bit, but I found that setting it to 100 gives a reasonable coverage.
Thank you so much, Kenneth! It worked perfectly for my needs. I’ll inform the Tomofast-x developers about this issue with the data and see if they can address it in a future pull request.
