controlling NaN pixels in ENVI scripting calls

Prev Go to previous topic

Go to next topic

Last Post 29 Aug 2011 12:35 PM by anon

3 Replies

Sort:


You are not authorized to post a reply.

Author

Messages

anon

New Member

Posts:

29 Aug 2011 12:35 PM
Is there an easy way, using ENVI programing calls, since I am working with imagery already opened with ENVI fids (vs IDL luns) to do this: (if any pixel in image = 0) then (pixel = NaN) ? Where the pixel locations vary, and therefore a mask would not work? I'd like to do something like this in a 'band math' call. Thank you...

MariM

Veteran Member

Posts:2396
Veteran Member

29 Aug 2011 01:10 PM
You could do a divide by 0 in band math: (float(b1) * (b1 ne 0)) / (float(b1) ne 0) Or write a function to use WHERE: FUNCTION find_nan, b1 indices = WHERE (b1 eq 0) ; find where there are 0 values b1 = float(b1) ; convert image to floating point if it is not already b1[indices] = !values.f_NaN ;set 0 values to NaN RETURN, b1 END

Deleted User

New Member

Posts:

29 Aug 2011 03:58 PM
Thank you, MariM; I did not know I could use 'where' in this sense!

Deleted User

New Member

Posts:

02 May 2012 10:08 AM
Thanks for this! I am working on 5-year time series MODIS data and it saved me from a lot of manual work. I am also still learning how to use IDL.

You are not authorized to post a reply.