PMC question

[R G W]

Hello sir,
I recently took your course on youtube and the course is very good.
After I finished the PMC section, I tried it with my own stp model, and I found that using triangulation for point judgment works well when the model I read is complete. But when the model I read is incomplete (such as half a sphere), the judgment of whether the point belongs to the inside of the model is not good (there are many points outside but considered to belong to the model), the problem is How to solve it?
The effect diagram is as follows.The model file is as follows.
before pmc:


after pmc:

 

[Quaoar]

Hello @R G W

This ray-casting PMC logic works for closed bodies exclusively, i.e., the boundaries should enclose a finite volume in space. Your model is an open shell, and that's why the classifier does not work properly. You have to cap your model to make it work. It's always like that when you use an intersection counter based classifier.

 

[R G W]

Hello @R G W

This ray-casting PMC logic works for closed bodies exclusively, i.e., the boundaries should enclose a finite volume in space. Your model is an open shell, and that's why the classifier does not work properly. You have to cap your model to make it work. It's always like that when you use an intersection counter based classifier.

Ok!Very thanks!