[GRASS-user] problems with colors resulting of an image fusion

[Moritz Lennert]

On 09/01/12 19:16, Moritz Lennert wrote:
> On 09/01/12 00:50, Markus Neteler wrote:
>> On Fri, Jan 6, 2012 at 5:42 PM, Moritz Lennert
>> <mlennert at club.worldonline.be> wrote:
>>> Hi everyone,
>>>
>>> I have some troubles with image fusion through IHS transformation of a
>>> Landsat image.
>>
>> Hi Moritz,
>>
>> could you please post the procedure you used?
>
> The same procedure as the one you describe in the book, i.e. the
> classical i.rgb.his & i.his.rgb approach with grey.eq color table for
> the resulting three bands.
>
> I'll try to write up a reproducible example, including relevant data if
> the NC set does not show the same behaviour.

Ok, here we go using nc_spm_08:

- Using "real-color" channels I get weird colors

g.region rast=lsat7_1002_20 at PERMANENT

r.colors lsat7_1002_30 at PERMANENT color=grey.eq
r.colors lsat7_1002_20 at PERMANENT color=grey.eq
r.colors lsat7_1002_10 at PERMANENT color=grey.eq

g.region -p rast=lsat7_1002_80 at PERMANENT

i.rgb.his red_input=lsat7_1002_30 at PERMANENT 
green_input=lsat7_1002_20 at PERMANENT blue_input=lsat7_1002_10 at PERMANENT 
hue_output=h intensity_output=i saturation_output=s --o

i.his.rgb hue_input=h at fusion_test 
intensity_input=lsat7_1002_80 at PERMANENT saturation_input=s at fusion_test 
red_output=fr green_output=fg blue_output=fb --o

r.colors map=fr at fusion_test color=grey.eq
r.colors map=fg at fusion_test color=grey.eq
r.colors map=fb at fusion_test color=grey.eq


Color composite before fusion (d.rgb -o red=lsat7_2002_30 at PERMANENT 
green=lsat7_2002_20 at PERMANENT blue=lsat7_2002_10 at PERMANENT):

http://geog-pc40.ulb.ac.be/grass/fusion_test/d_rgb_30_20_10_grey_eq.png

Color composite after fusion (d.rgb -o red=fr at fusion_test 
green=fg at fusion_test blue=fb at fusion_test):

http://geog-pc40.ulb.ac.be/grass/fusion_test/d_rgb_fr_fg_fb_grey_eq.png


- Using near-infrared false color composition the colors are as expected:


g.region rast=lsat7_2002_30 at PERMANENT

r.colors lsat7_2002_40 at PERMANENT color=grey.eq
r.colors lsat7_2002_30 at PERMANENT color=grey.eq
r.colors lsat7_2002_20 at PERMANENT color=grey.eq

g.region -p rast=lsat7_2002_80 at PERMANENT

i.rgb.his red_input=lsat7_2002_40 at PERMANENT 
green_input=lsat7_2002_30 at PERMANENT blue_input=lsat7_2002_20 at PERMANENT 
hue_output=h intensity_output=i saturation_output=s --o

i.his.rgb hue_input=h at fusion_test 
intensity_input=lsat7_2002_80 at PERMANENT saturation_input=s at fusion_test 
red_output=pir_fr green_output=pir_fg blue_output=pir_fb --o

r.colors map=pir_fr at fusion_test color=grey.eq
r.colors map=pir_fg at fusion_test color=grey.eq
r.colors map=pir_fb at fusion_test color=grey.eq

Color composite before fusion (d.rgb -o red=lsat7_2002_40 at PERMANENT 
green=lsat7_2002_30 at PERMANENT blue=lsat7_2002_20 at PERMANENT):

http://geog-pc40.ulb.ac.be/grass/fusion_test/d_rgb_40_30_20_grey_eq.png

Color composite after fusion (d.rgb -o red=pir_fr at fusion_test 
green=pir_fg at fusion_test blue=pir_fb at fusion_test):

http://geog-pc40.ulb.ac.be/grass/fusion_test/d_rgb_pir_fr_fg_fb_grey_eq.png

Can anyone reproduce this ? Any explanation ?

Moritz