On Wed, 14 Apr 2010 15:47:01 -0400, Prof. Jeffrey Cardille wrote: > For anyone > interested, I am trying to repair Landsat satellite photos, which have big > broken strips of nodata due to a mechanical failure. The strips are up to 14 > pixels wide. I need to do this 2000x2000 interpolation about 2000 times-- so > speed differences of even a few seconds or minutes are quite important..
Hello, I'm a graduate student at the University of Minnesota in the GIS program. For the past couple years I have been assisting a conservation biology doctoral student with GIS aspects of her dissertation, which involved acquiring Landsat data. I presume you are talking about the scan line corrector problem that developed in 2003 on Landsat 7 in the extended thematic mapper? USGS engineers at EROS have tried several approaches to correcting the gaps, primarily by interpolation and filling in from other scenes from the same season. I don't have a reference handy, but I did read something that found that method had an unreasonable amount of cloud cover interference. However, the EROS folks eventually published a paper describing how their "multi-scale segmentation" technique worked that was then used for EROS terrain-corrected products. If you haven't read it, the paper has lots of valuable information on the exact nature of the problem. Here's a link to the citation from the ACM portal: http://portal.acm.org/citation.cfm?id=1393487 For my purposes, the EROS corrected data were sufficient, but some applications such as permeability studies may require pixel-level precision. Good luck with your endeavor and if you find a solution, I hope you'll post about it here. Thomas Juntunen MGIS Program, University of Minnesota _______________________________________________ R-sig-Geo mailing list R-sig-Geo@stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/r-sig-geo
