Skip to main content
Solved

Generalize building and keep general shape?


aron
Supporter
Forum|alt.badge.img+16
  • Supporter

I am looking for a method to remove the outcrops, circled in blue, from the building fotprints below. The generalization algorithms I have tried "soften" the polygons to much and make them look a bit organic. I would like to keep 90 degree corners as much as possible so that the viewer quickly can identity the shape as a building.

It might require a different process but I would also greatly appreciate suggestions on how to fill out the "jagged" corners circled in red.

 

Cheers!

 

Building2

Best answer by caracadrian

How about this:

Chop into lines, extend them, build topology then areas and keep the largest area for each feature.

You have to figure out a value for line extension that either generally applies to your data or as a derivate of the shape/number of vertices/segment length.Building generalization idea

View original
Did this help you find an answer to your question?

4 replies

caracadrian
Contributor
Forum|alt.badge.img+23
  • Contributor
  • October 20, 2021

Have you tried Deveau algorithm in Generalizer?


aron
Supporter
Forum|alt.badge.img+16
  • Author
  • Supporter
  • October 20, 2021
caracadrian wrote:

Have you tried Deveau algorithm in Generalizer?

Yes, Deveau does remove the outcrops, but does not retain she square shape of the building. Ie corners don't remain at 90 degrees. At least, I have not found a way to keep walls aligned.


caracadrian
Contributor
Forum|alt.badge.img+23
  • Contributor
  • Best Answer
  • October 20, 2021

How about this:

Chop into lines, extend them, build topology then areas and keep the largest area for each feature.

You have to figure out a value for line extension that either generally applies to your data or as a derivate of the shape/number of vertices/segment length.Building generalization idea


aron
Supporter
Forum|alt.badge.img+16
  • Author
  • Supporter
  • October 20, 2021
caracadrian wrote:

How about this:

Chop into lines, extend them, build topology then areas and keep the largest area for each feature.

You have to figure out a value for line extension that either generally applies to your data or as a derivate of the shape/number of vertices/segment length.Building generalization idea

Thanks! That is a really good idea! Will try it out.


Cookie policy

We use cookies to enhance and personalize your experience. If you accept you agree to our full cookie policy. Learn more about our cookies.

 
Cookie settings