Subject: CGAL users discussion list
List archive
- From: diethelm <>
- To:
- Subject: [cgal-discuss] fast algorithm for special case of 2D Minkowski sums?
- Date: Thu, 3 Jan 2013 02:59:30 -0800 (PST)
Dear all,
I hope this forum is the right place to ask my question. If not, I would
appreciate any pointers to the correct place.
My problem is that I want to compute Minkowski sums of two simple but
non-convex polygons P and Q, say, with m and n vertices, respectively. The
polygons are actually simply connected, i.e. there are no holes. Classical
knowledge tells me that the usual algorithms solve this problem in O((mn)^2)
time. Now my particular interest is in the two special cases where P = Q and
P = -Q. Thus, a straightfoward application of the standard algorithm gives a
method that has an O(n^4) complexity. Is there an algorithm for these two
special cases, or at least for one of them, that has a lower complexity?
Many thanks for your help.
Kai Diethelm
--
View this message in context:
http://cgal-discuss.949826.n4.nabble.com/fast-algorithm-for-special-case-of-2D-Minkowski-sums-tp4656437.html
Sent from the cgal-discuss mailing list archive at Nabble.com.
- [cgal-discuss] fast algorithm for special case of 2D Minkowski sums?, diethelm, 01/03/2013
- Re: [cgal-discuss] fast algorithm for special case of 2D Minkowski sums?, Efi Fogel, 01/03/2013
Archive powered by MHonArc 2.6.18.