Skip to Content.
Sympa Menu

cgal-discuss - Re: [cgal-discuss] lower dimensional features of minkowski sum?

Subject: CGAL users discussion list

List archive

Re: [cgal-discuss] lower dimensional features of minkowski sum?


Chronological Thread 
  • From: Efi Fogel <>
  • To:
  • Subject: Re: [cgal-discuss] lower dimensional features of minkowski sum?
  • Date: Sun, 15 Jul 2007 14:57:35 +0300

Michael George wrote:
Hi,

For my application I would like to be able to find points on the boundary of the minkowski sum of the interior of some polygons, but it looks like the new code for minkowski sums drops the lower-order features. I would also like to take sums of objects that are composed of lines and circular arcs, although that's not as important to me since I can at least sum polygons and circles.

Can anyone advise me on whether CGAL does in fact support this, whether there are plans for it in the future, and whether it might be feasible for me to modify the existing code to support this?
Regarding the lower-order features, I assume that it is possible to exploit the existing code. It means that the input operands may have lower-order features, and that they must be represented somehow. One solution is as follows:
1. Identify the lower-order features, remove them, and obtain simple polygons.
2. Compute the Mink. sum of the simple polygons using the Minkowski_sum_2 package
3. Compute the Mink. sum of pairwise point sets that have lower-order features. This is a relatively easy task, and results with points sets the boundary of which may have lower-order features
4. Identify the lower-order features, remove them, and obtain simple polygons.
5. Compute the union of the results of 2. and 4 using the Boolean_set_operation_2 package
6. Compute the union of the results of 5 and lower-order features identified at 4.

Alternatively, instead of steps 4., 5., and 6., you might be able to use Nef_2.

Indeed, the code supports computations of Minkowski sum of (linear) polygons and of a (linear) polygon and a (whole) disc. At this point there are no plans to enhance the code to supports point sets the boundaries of which are different than the above.

In general, our aim was to stay within the (exact) rational arithmetic realm. The Minkowski sum of point sets the boundaries of which have circular arcs might be something that cannot be computed and represented using rational arithmetic. Let us know in case you decide to modify the code yourself and need further help.
Thanks!

Mike
--
____ _ ____ _
/_____/_) o /__________ __ //
(____ ( ( ( (_/ (_/-(-'_(/
_/






Archive powered by MHonArc 2.6.16.

Top of Page