Topic: Correlation of two satellite images


I am just writing my bachelor thesis about dune migration in Qatar. I am doing most of my work with ArcGIS but I also would like to write about the possibility to use COSI-Corr to analyse dune migration. For that I would like to show an simple analysis with COSI-Corr.

I would like to use the picture down there, which is from 2004 and compare it to almost the same picture just from 1995.

Both pictures are from qatari georeferenced satellite data. They are cut out with a rectangular masc in ArcGIS and have the same resolution of one meter. They are saved as tif-data. The only difference is that the picture of 1995 is black and white and just has one band, while the same picture of 2004 has three bands. I tried straight away to analyse the pictures with "COSI-Corr - Correlation" and used a lot of different options sometimes runing through and sometimes not. But the only results I got are black and white images showing pixels.

Now I would like to ask if anyone knows the mistake I am doing? I have the feeling the pictures are ready to work with, but as I have not much ENVI-experience maybe I am missing something out. I could upload the second picture if it helps or a screenshot of the results?

Thank you for reading and helping

Best regards, Fabian

2 (edited by fabian.boe 2013-05-13 08:02:51)

Re: Correlation of two satellite images


The size of the original images is 1303x1004 Pixel



Re: Correlation of two satellite images

What parameters did you use in the correlation? And I suppose this post should go in the correlation section...