Here's another issue I came across, and I do not have any solution so far. What I'm doing is sampling CAD faces in their UV spaces with uniform grids. Something like that:
![1628062119912.png 1628062119912.png](http://analysissitus.org/forum/data/attachments/0/37-5053bf5cf9463ce6696f28feb4cbc340.jpg)
And with the grid overlay:
![1628062174299.png 1628062174299.png](http://analysissitus.org/forum/data/attachments/0/38-a46e4fe1f71e8310bb17d4735e055aa9.jpg)
The red color means that the corresponding point falls inside the parametric domain of a face. The white color means that the corresponding point is outside. For that test, I use the good old
Has anyone ever implemented fast PMC for 2D contours?
![1628062119912.png 1628062119912.png](http://analysissitus.org/forum/data/attachments/0/37-5053bf5cf9463ce6696f28feb4cbc340.jpg)
And with the grid overlay:
![1628062174299.png 1628062174299.png](http://analysissitus.org/forum/data/attachments/0/38-a46e4fe1f71e8310bb17d4735e055aa9.jpg)
The red color means that the corresponding point falls inside the parametric domain of a face. The white color means that the corresponding point is outside. For that test, I use the good old
IntTools_FClass2d
and it appears to be not that good (just old). The problem is the performance and now I'm looking for a better way of point-face classification.Has anyone ever implemented fast PMC for 2D contours?