Point cloud pre-processing

[VeeranjaneyuluToka]

I was looking into the both paper and implementation and got a few comments or clarifications.

-> Why are you using Normal estimation, Inensity normalization and segmentation together? I mean i try to understand the dependency on each other.

-> why are we using conventional/un-supervised based segmentation in case PCLs? Can not we use any of the recent segmentation models similarly the way SAM is being used in case of image?

Thanks, Veeru.