Some formula +++The algorithm's steps are: 1. Read pixel sizes from Image files and calculate R, the ratio of multispectral cell size to high-resolution cell size 2. Apply the High-pass filter to the high spatial resolution image. 3. Resample the Multi-Spectral image to the pixel size of the high-pass image. Note, bilinear resampling required (4 nearest neighbours)! 4. Add the HPF image weighted relative to the global standard deviation of the Multi-Spectral band. 5. Optionally, stretch linearly the new HPF-Sharpened image to match the mean and standard deviation of the input Multi-Sectral image.
g.region rast= -p # R, G, B composite at 28.5m d.rgb b=lsat7_2002_10 g=lsat7_2002_20 r=lsat7_2002_30 # Brovey fusion i.fusion.brovey -l ms1=lsat7_2002_20 ms2=lsat7_2002_40 \ ms3=lsat7_2002_50 pan=lsat7_2002_80 \ outputprefix=brovey # display at 14.25m g.region rast=brovey.blue -p d.rgb b=brovey.blue g=brovey.green r=brovey.red
![]() R, G, B composite of Landsat at 28.5m |
![]() R, G, B composite of Landsat Brovey fusion at 14.25m |
Last changed: $Date: 2011-11-08 11:42:51 +0200 (Tue, 08 Nov 2011) $