11
notroot
208d

# I'm obsessed with convex hulls.

• 4
What size?
• 6
Delaunay triangulation!
• 4
This sounds like "I like big butts and I cannot lie".
• 1
What size?
• 5
@dorapovah @Wisecrack is this also you ?
(If yes, there's absolutely no need to create multiple accounts to get your point across. I saw it once. Would've replied without the second.)
(If not, then nevermind.)

It's not about the size of the hull but the concept and also the methods we use to find it.

I'm currently just obsessing over hull merge and the data structure used to represent convex hulls mainly the half edge representation. Preparata's original paper and the one later with Amato are great. In the latter one, they outlined the method to find seam edges. That's a really really good explanation of the concept. Most articles or online notes are crap or incomplete.
• 2
@-red Thats not me. I think they just had the same question lol.

Convex hull mergers. Hmm. I wonder how I can relate this back to prime factorization.

If a prime could be represented as a convex hull, and a product could be represented as the merger of two convex hulls, perhaps there is a geometric solution in this representation, for fast prime factorization?

Are there any algorithms for "de-merging" convex hulls?
• 1
• 3
@Wisecrack Ah, never mind then. Sorry!

For splitting, same algorithms as CH can be used on the hanging edges after the connecting edges have been severed. It's fundamentally the same problem as finding CH.

I'm more in awe of CGAL than I was before. Earlier, I mostly just used it for my point cloud processing. But now I have a whole new appreciation for it's implementation.
• 2
@Wisecrack I meant to say exposed edges, not hanging (they don't exist anymore).
• 2
@-red what's CGAL? If it's in the papers, know that I didnt read them. Just did a cursory review of the first pages to see if they might be relevant to you.
• 4
@Wisecrack It's a C++ library for Computational Geometry. I love it.

(Computational Geometry Algorithms Library)

It's got a bit of a learning curve but it's great once you do. I stumbled upon it sometime around twenty eighteen when I needed to do some line of sight/visibility computations on large point clouds. Slightly slow to compile but it runs fast!!
• 2
Wow. @dorapovah hasn't said anything since.

I'm so curious. That was the only thing @dorapovah wanted to know. That seemed to be the purpose of @dorapovah.
• 0
@-red what's the line of sight algorithm being used for?

What are you working on (if you dont mind my asking)?