Take the 2-minute tour ×
Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. It's 100% free, no registration required.

I am editing a single-band .tiff file in ArcGIS 9.3. I started with an image that I had to georeference, but unfortunately, the image had these little slivers of missing data between the data cells. I get left with a raster file that has lots of "slivers" of nodata between my coding cells (shown here in white). enter image description here

I would like to use something analogous to vector snapping, in which I eliminate those white slivers, instead making those sliver raster cells equal to the value of one of the adjacent cells. thank you.

share|improve this question
    
I should note that I am OK with increasing the raster cell size slightly and loosing a bit of resolution; as long as it gets rid of those missing data areas. –  Luke Nov 5 '12 at 22:33
    
Also of note is that I have a raster layer that contains just the sliver cells. –  Luke Nov 5 '12 at 22:34
add comment

2 Answers

Based on the image you've provided, it looks like this might be a raster that was originally relatively coarse that was resampled to a much higher resolution. I'm not sure exactly how the no data slivers occurred, but you might try resampling it back to a coarser resolution (try to approximate what the minimum cell size should be by visually inspecting your raster). Choose Majority for your resampling technique.

Edit: I just recalled that raster processing in Arc ignores no data, so you'll have to give the no data areas a unique value before you resample. Use the Con tool with IsNull to set the no data areas to a unique value of your choosing.

share|improve this answer
    
That's right egdetti, I am trying to salvage a dataset that was originally created with Arc, Exported as an image, then the original raster was lost. The nodata slivers were originally some kind of intermediate borders between solid colors that the image editing software applied - I deleted them, leaving the slivers. I'll see if I can resample back to the original resolution (30 m Landsat pixels). Thanks for the tip. –  Luke Nov 6 '12 at 2:44
add comment
up vote 2 down vote accepted

After banging my head on the wall for a while, I figured it out. First, I used reclassify in spatial analyst to change the values for the slivers to a unique value (99). Then I used Hawth's Tools' raster tool called "spatial replace" to replace values of 99 with the value representing the mean of the neighborhood.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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