I haven't come across the notion of "penetration depth" before - very
interesting! It seems that the usual definition is:
The (overall) penetration depth, (A;B) is defined as the minimum amount
of translational motion required in any direction required to make the
interior of A disjoint from B
(from this paper:
http://www.cs.duke.edu/courses/spring07/cps296.2/course_projects/shashi_proj.pdf
)
It looks like this is a non-trivial computation, except possibly in the
case of both polygons being convex. There's nothing in JTS at the
moment which computes this value. The technique used by the distance
algorithm to compute distance = 0 is quite different and much simpler,
and can't be amended to compute penetration depth (As you might see,
there's a big difference between computing a 0 value for something, and
computing a non-zero value which describes a complex situation)
The survey paper lists the current best algorithms for various cases
(both convex, one convex). I'm not sure if they're similar to the GJK
algorithm. Seems like they'd be fairly complex to implement (unless
perhaps some existing code can be found and adapted to JTS).
Jim Kay wrote:
Is there an easy way to calculate by how much geometries (say
polygons) overlap?
I am thinking of the collision-detection problem of 'penetration depth'.
The distance method returns zero for overlapping polygons - could this
be amended to give a negative distance?
Would my request require something like the GJK algorithm, and would
that be difficult to implement?
*/Jim/*
Jim Kay
Senior Engineer
Interfleet Technology Ltd
email: ka...@interfleet.co.uk
tel: +44 (0) 1332 223229
fax: +44 (0) 1332 223181
mobile: +44 (0) 7715 536828
web: _www.interfleet.co.uk_ <file://www.interfleet.co.uk>
Interfleet Technology Ltd
Interfleet House
Pride Parkway
Derby
DE24 8HX
**********************************************************************
Interfleet Technology Ltd
Interfleet Technology's email disclaimer is found via the following
URL: http://www.interfleet-technology.com/disclaimer.asp
Registered Offices:
http://www.interfleet-technology.com/offices.asp
Please consider the environment before printing this e-mail.
**********************************************************************
------------------------------------------------------------------------
_______________________________________________
jts-devel mailing list
jts-devel@lists.jump-project.org
http://lists.refractions.net/mailman/listinfo/jts-devel
--
Martin Davis
Senior Technical Architect
Refractions Research, Inc.
(250) 383-3022
_______________________________________________
jts-devel mailing list
jts-devel@lists.jump-project.org
http://lists.refractions.net/mailman/listinfo/jts-devel
