[uf-discuss] Geographic polygons [was geo - accuracy of coordinates

[Kevin Marks]

On Oct 3, 2006, at 1:12 AM, Andy Mabbett wrote:

> In message <FB2D8589-4A17-487D-8C8A-B4BCE0BC37B3 at placenamehere.com>,
> Chris Casciano <chris at placenamehere.com> writes
>
>> Totally ignored the point I was trying to make... and that is that
>> describing a border - of any shape - by the use of a collection of  
>> geo
>> coords (at whatever precision) is a totally different task then
>> defining an individual point and its precision.
>
> Totally ignored the point I was trying to make... and that is that it
> would perhaps be better to have the capacity to describe a polygon.

They aren't mutually exclusive, which is why I suggested separating 
polygons into a separate thread. There is an HTML way to express 2D 
polygons in image maps, so what we would need is a way to georeference 
the imagemaps to translate these back to earth-based co-ordinates.

For imagery in lat-long space, or for close-in zooms, just specifying 
the lat/long of the corners would be adequate; for other projections 
such as Peterson, Roberts or Mercator, you may need to specify the 
transform with more care, if trapezoidal interpolation give significant 
distortion.