comp.lang.idl-pvwave archive: archive » Understanding IDLanROI

Home » Public Forums » archive » Understanding IDLanROI

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: Understanding IDLanROI [message #73088 is a reply to message #72990]

Fri, 22 October 2010 07:16

KRDean
Messages: 69
Registered: July 2006

Member

On Oct 20, 11:01 am, Guillermo
<guillermo.castilla.castell...@gmail.com> wrote:
> kBob writes:
>> I am attempting to use IDLanROI to determine if Shapefiles overlap
>> another Shapefile...
>> When I tried to use the US shapefile in IDLanROI (and IDLgrROI) there
>> was a line running across the United States from the Northeast to
>> Hawaii.
>
> I think what happened is that when you created those ROI objects, you
> provided the entire set of vertices that define the shape of the US as
> a single part, and then the last vertex of the ring defining the
> conterminous US got connected with the 1st vertex in Hawaii, hence the
> line you mention. I think the IDLanROIGroup::ContainsPoints method
> should work properly if you create the corresponding object by adding
> sequentially the parts in the US shapefile entity, providing neither
> of the parts are holes (e.g., imagine that Colorado had successfully
> opted out of the Union; the ring defining its shape would be a hole,
> and its vertices would be listed counterclockwise in the shapefile
> entity). Since afaik there are no holes in the US territory, the above
> method should work without having to resort to the tessellator or
> other graphic object classes.
>
> But in a more general application (i.e., where there are polygons with
> holes), I wonder what would be the way to find out whether a point is
> inside a polygon that contain holes. The IDLanROIGroup cannot be used
> for that, as there is no way to specify that a given ROI within the
> groups is actually a hole. The easiest solution would be that the
> ITTVIS folks implement the IDLffShape::ContainsPoints method. Now that
> there is a larger interest to integrate image analysis with vector
> GIS, there is some hope that this will happen in a future release...
>
> Guillermo

Your right, the Tessellator is not required. I just have to perform
the ContainsPoints() on the individual shapefile parts.

To find the polygons that are holes, I use the IDL routine POLY_AREA.
The /SIGNED keyword returns the area as either positive or negative,
which tells you if the vertices are counterclockwise or clockwise. All
the State polygons from states.shp are clockwise.

Kelly

Report message to a moderator

[Message index]

		Understanding IDLanROI By: KRDean on Fri, 08 October 2010 14:13
		Re: Understanding IDLanROI By: KRDean on Tue, 19 October 2010 08:02
		Re: Understanding IDLanROI By: David Fanning on Tue, 19 October 2010 10:38
		Re: Understanding IDLanROI By: guillermo.castilla.ca on Wed, 20 October 2010 10:01
		Re: Understanding IDLanROI By: KRDean on Fri, 22 October 2010 07:16
		Re: Understanding IDLanROI By: Matt[2] on Mon, 25 October 2010 09:17
		Re: Understanding IDLanROI By: guillermo.castilla.ca on Wed, 27 October 2010 08:44
		Re: Understanding IDLanROI By: guillermo.castilla.ca on Mon, 01 November 2010 06:23
		Re: Understanding IDLanROI By: guillermo.castilla.ca on Mon, 01 November 2010 06:24
		Re: Understanding IDLanROI By: David Fanning on Tue, 19 October 2010 14:02
		Re: Understanding IDLanROI By: KRDean on Tue, 19 October 2010 13:19

Previous Topic:	Re: Still missing features in IDL 8
Next Topic:	Subtracting arrays of same dimension, different size - interpolating one to fit the other

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Oct 10 06:35:24 PDT 2025

Total time taken to generate the page: 0.47510 seconds