Greetings!
I am using Landsat 7 ETM+ SLC-Off imageries for my academic project work. As users may be aware due to Scan line corrector going off in the satellite, large gaps are created in each band of image. It accounts for loss of 20 to 22 percent of data. Since recently, gap mask files are provided along with Landsat data, so you can tell where actual data is versus the gaps, also all data gaps have a digital value of 0 (zero).
Options to use fill in those gaps created include using interpolation methods, usage of images from different dates or sensors like IRS or MODIS, usage of geostatistics etc. I am interested to know if anyone has employed such methods in GRASS previously. Proven techniques would help me progress faster.
Thanks and Regards,
Chethan S.
–
Chethan S.
On Friday 03 of December 2010 16:35:21 Chethan S wrote:
I am using Landsat 7 ETM+ SLC-Off imageries for my academic project work.
As users may be aware due to Scan line corrector going off in the
satellite, large gaps are created in each band of image. It accounts for
loss of 20 to 22 percent of data. Since recently, gap mask files are
provided along with Landsat data, so you can tell where actual data is
versus the gaps, also all data gaps have a digital value of 0 (zero).
Options to use fill in those gaps created include using interpolation
methods, usage of images from different dates or sensors like IRS or MODIS,
usage of geostatistics etc. I am interested to know if anyone has employed
such methods in GRASS previously. Proven techniques would help me progress
faster.
I used two or three SLC-Off images to create a composite. Rather a
straightforward job using r.patch. Make sure you understand how r.patch works
(the order of the supplied raster maps matters). Of course, there is no doubt
that you can play with the various interpolation techniques. In this case I
did not so I can't give more info on it.
Nikos
Hi Chetan,
I have been working a lot with SLC-Off imagery lately. Some people in my department have used the gap filling programs floating around the net, but I’m not familiar with them personally. I’ve settled on using r.patch as well, although you have to be careful how you apply it. I found that even if I used radiometrically corrected landsat images (using i.landsat.toar) from the same season often the patched parts of the image did not fit smoothly with the rest of the image (i.e. you could see striations where the gaps had been). I’m using the imagery for land classification, so I’ve found it works better if I do the classification on each landsat image separately and then patch them. Using this method you can’t tell where the former gaps are, at least in my experience working with imagery from the dry season in southeast asia.
Good luck!
Nick
On Fri, Dec 3, 2010 at 10:35 PM, Chethan S <chethanuniversal@gmail.com> wrote:
Greetings!
I am using Landsat 7 ETM+ SLC-Off imageries for my academic project work. As users may be aware due to Scan line corrector going off in the satellite, large gaps are created in each band of image. It accounts for loss of 20 to 22 percent of data. Since recently, gap mask files are provided along with Landsat data, so you can tell where actual data is versus the gaps, also all data gaps have a digital value of 0 (zero).
Options to use fill in those gaps created include using interpolation methods, usage of images from different dates or sensors like IRS or MODIS, usage of geostatistics etc. I am interested to know if anyone has employed such methods in GRASS previously. Proven techniques would help me progress faster.
Thanks and Regards,
Chethan S.
–
Chethan S.
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user
Agreed. You have to make sure you have valid training areas are for each image you want to classify. I’ve also found that dealing with clouds and their shadows is one of the biggest issues to consider when doing classifications.
On Sun, Dec 5, 2010 at 4:12 AM, Nikos Alexandris <nikos.alexandris@uranus.uni-freiburg.de> wrote:
Nick Jachowski:
I have been working a lot with SLC-Off imagery lately. Some people in my
department have used the gap filling programs floating around the net, but
I’m not familiar with them personally. I’ve settled on using r.patch as
well, although you have to be careful how you apply it. I found that even
if I used radiometrically corrected landsat images (using i.landsat.toar)
from the same season often the patched parts of the image did not fit
smoothly with the rest of the image (i.e. you could see striations where
the gaps had been).
Right. The same here. But I used the composites only for visual interpretation
which was ok. It’s always interesting how different tasks pose different
challenges.
I’m using the imagery for land classification, so
I’ve found it works better if I do the classification on each landsat
image separately and then patch them. Using this method you can’t tell
where the former gaps are, at least in my experience working with imagery
from the dry season in southeast asia.
Interesting. Yet, I guess, you had to use independent training areas (in case
you performed supervised classifications), right?
[…]
Nikos A