| 12 Jun 2013 05:47 AM |
|
|
Hello ENVI Users,
i have a question concerning the mosaicking tool. I did mosaicking for several Landsat TM Scenes and compared the results with and without color balancing. (I used some transects to analyse the spectral shift at the border of the scenes.)
At most points the shift in spectral signature was higher with color balancing, and lower without color balancing (so the opposite of how it should be).
Is there any explanation for this effect?
Many thanks in advance
|
|
|
|
Deleted User
New Member
Posts:  
| 12 Jun 2013 09:10 AM |
|
|
Can you provide an example of what you mean by 'shift in spectral signature'. Color balancing just adjusts the RGB values but not the actual DN values.
|
|
|
|
Deleted User
New Member
Posts:
| 13 Jun 2013 02:48 AM |
|
|
Thanks for your reply,
I actually mean the gray level jump at the border between the mosaicked images. It doesn't matter which point of the border i consider or which channel i use. The gray level jump seems to be higher with color balancing in most cases.
I thought that using 'color balancing' has the same effect as any kind of Histogram Matching in other programs,namely the adjusting of the gray level values of the one image to the histogram of another image (e.g. to prevent a false classification as a result of the grey level jump)
I am using calibrated an atmosphere-corrected scenes.
|
|
|
|
Deleted User
New Member
Posts:
| 17 Jun 2013 01:22 AM |
|
|
Does anybody have some hints fo me?
|
|
|
|
Deleted User
New Member
Posts:
| 17 Jun 2013 11:52 AM |
|
|
You could try using 'edge feathering' or 'cutline feathering' to see if your results get better.
|
|
|
|