1

enter image description here

I am using ArcGIS Desktop 10.5.1.

I have a raster image which I have extracted from a large raster data, and ended up having some scattered small raster groups on the extracted image. This scattered raster are very difficult to deal with when changing the raster to polygon.

How can I delete scattered raster data sets which are of no use?

  • 1
    Can you share a screenprint of the issue? Which software are you using? Are continuous or discrete values? – aldo_tapia Nov 2 '17 at 1:53
  • So the image that I just added is a zoomed to layer. You can see that the main raster is swayed to the right. The software is 10.5.1. – Sim Nov 2 '17 at 2:01
  • I believe the data set is continuous. i have encircled the scattered sets with red – Sim Nov 2 '17 at 2:02
  • What software, not just what version. ArcGIS? – BradHards Nov 2 '17 at 2:53
  • Did you use a polygon to extract/clip this image? – ahmadhanb Nov 2 '17 at 2:59
2

Your question is unclear when it says delete them out. Perhaps you mean to reclassify them to NODATA but retain the extent of the raster or clip back the raster. You need to be providing more information.

Assuming the white represents NODATA and the large area in the top right is the only bit of the raster you want to keep then simply clip it out. This would significantly reduce the file size.

  • I think if I reclassify them to no data according to their pixel values, then I will lose similar data sets with same pixel values. I agree with clipping, but I think I will need a clipping poygon to use, which i dont have – Sim Nov 2 '17 at 21:43
  • The white represents no data. By losing similar data sets with same pixel values, I mean losing data with similar pixel values (which is elevation in this case) from my raster to the right. – Sim Nov 2 '17 at 22:34
  • If you follow the link I provided to the Clip tool and study the help file you will see you do not need a clipping polygon you simply define the rectangle based upon the coordinates which you can easily determine from your dataset. – Hornbydd Nov 2 '17 at 23:51
  • gis.stackexchange.com/users/934/hornbydd, Thank you very much!! I really owe you a gratitude.It worked well – Sim Nov 3 '17 at 1:30

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.