Another aproach could be to compare angles in each vertex "corner". If all angles are the same in some sort of tolerance the shape should be the same.
Jusr a thought /Nicklas -- Sendt fra min Nokia N9 09.05.12 18:41 skrev Bob Pawley: Hi Sandro Absolutely, if it's something people want. I developed the first part a few months ago, just as something to do . I compare relative distance of points(intersections) of the object, overriding the actual distances(size) and ignoring any spatial reference (placement). The result is an array set of numbers that should match if the two objects are the same. This system can be easily scaled to improve resolution. At the moment I use a resolution of 100 points. I still need to develop a method of analyzing the arrays ( and perhaps, giving a percent evaluation of sameness) for which I have some ideas. I will probably need some help in making it an actual Postgis function if that is what you have in mind. Bob -----Original Message----- From: Sandro Santilli Sent: Wednesday, May 09, 2012 9:13 AM To: PostGIS Users Discussion Subject: Re: [postgis-users] Same, similar et al On Wed, May 09, 2012 at 09:06:15AM -0700, Bob Pawley wrote: > Hi Steve > > FYI - I'm not involved in the geography side of Postgis, I'm > developing an application to analyze engineering drawings. So, same > in shape even if the shapes are spatially diverse is what I was > looking for. > > However, since this function doesn't exist I will need to develop > something. Will you share it ? --strk; _______________________________________________ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users _______________________________________________ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users
_______________________________________________ postgis-users mailing list postgis-users@postgis.refractions.net http://postgis.refractions.net/mailman/listinfo/postgis-users
