Follow-up question about FLAASH and ASTER L1B data

Prev Go to previous topic

Go to next topic

Last Post 11 Oct 2013 04:40 AM by anon

1 Replies

Sort:


You are not authorized to post a reply.

Author

Messages

anon

New Member

Posts:

11 Oct 2013 04:40 AM
Hi again, forum-- One more question about FLAASH with ASTER L1B data. The tutorial suggests that there should be a SCENECENTER attribute in the metadata, but there is none--only 4 GRINGPOINT entries each for latitude and longitude. It also says that I need to stack/rotate the image first, then get the center point from there---this part is fine--do I need to mask out the no-data (black) strips on the edge first? in other words, is the workflow: 1. Stack VNIR/SWIR bands 2. Rotate image 3. trim the no data sections 4. calculate the raster center (in ENVI or a GIS program or whatever) Lot of busy-work required to prepare these images, so I want to make sure I'm doing it correctly! Thanks-- Steve

MariM

Veteran Member

Posts:2396
Veteran Member

11 Oct 2013 07:03 AM
The tutorial is specific to the 1A dataset so some steps (such as georeferencing or orthorectifying) are not necessary. You do want to stack the bands to get the best result from FLAASH because it performs the correction on the spectra (pixel-by-pixel). When you stack the images, the image will automatically be rotated to north up, so there is not an extra step there either. Masking is not necessary for FLAASH because it processes the data pixel-by-pixel but it would be a required input for Quac which uses the entire scene statistics. The 'scene center' does not have to be that specific. Just use your cursor location tool and get a location of the approximate scene center.

You are not authorized to post a reply.