I've used Simplify Polygon with 'Maximum Allowable Offset' 250m, and then Aggregate Polygons with 'Aggregation Distance' also 250m (opposite order also tested). Still I have thin parts of polygons as below: enter image description here

I presume that I need some minimum angle tool. But there are also thin inner parts like this: enter image description here Yes, the width is only 35m, while I want objects have at least 250m. In this case propably 'Split Narrow Points' would be helpful, but I don't have this tool. Any ideas please?

  • There are several simplify options; which one did you use? The Zhou-Jones method might give you better results than the default Douglan_Peuker. – johns Dec 6 '18 at 13:15
  • I can see only 'POINT_REMOVE' and 'BEND_SIMPLIFY' algorhitms :( – D.J. Mac. Dec 6 '18 at 13:29

You can buffer with a negative buffer distance, for example -25 then buffer again with +25.

Another option is Integrate:

Integrate is used to maintain the integrity of shared feature boundaries by making features coincident if they fall within the specified x,y tolerance. Features that fall within the specified x,y tolerance are considered identical or coincident.

Be sure to backup your data Before you try it since it will modify the input. Set cluster tolerance as small as possible:

enter image description here

|improve this answer|||||
  • 1
    Thank you so much :) Result is very similiar to expected one :) – D.J. Mac. Dec 6 '18 at 13:15
  • My upvote was for the negative buffer followed by a positive buffer. This method typically does the best job of maintaining the original geometry where desired, though it will slightly round off any concave corners. It's the quickest, easiest, and most intuitive way to eliminate narrow peninuslae and isthmuses. – Tom Dec 6 '18 at 17:14
  • @D.J.Mac. please accept my answer with the checkbox if it solved your problem – BERA Dec 7 '18 at 14:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.