[GRASSLIST:2338] Re: i.his.rgb

[Markus Neteler]

On Thu, Jan 22, 2004 at 12:22:20PM -0800, Ian Macmillan wrote:
> Hi all, I have a problem that I can't quite figure out with i.his.rgb.  I have
> been manipulating landsat 7 data using the method described in the GRASS book. 
> I am trying to add in band 8.  So far I can get this to work, in a number of
> different ways.  The way I have found to be the most successful (for retaining
> the colors that I want) is to enhance the contrast of all 4 bands using
> r.rescale, converting to his (i.rgb.his), and converting back to rgb (i.his.rgb
> with band 8 as intensity).  The problem is that a lot of "holes" appear in the
> final rgb files.

I don't think that the usage of r.rescale is a good idea. Like that
you are manipulating the pixels instead of the color table.

Why not applying the HIS/RGB procedure and then using r.colors on
the resulting 3 channels?

>  Single pixels seem to be converted to a value of zero, even
> though the same pixels (in space) are not zeros or nulls in the original files
> (pre i.rgb.his > i.his.rgb).  I have posted a specific example here.  

Yes - probably due to r.rescale usage.

> www.uweb.ucsb.edu/~ian_macmillan/hisrgb.jpg
> 
> Band1x is the i.his.rgb converted blue file, while bands 1,2,3,8 are the
> original rescaled files.  The black dots are clearly visible scattered
> throughout the image.  They each have a value of zero.
> 
> Anybody have any ideas what is going on?

It's not the procedure proposed in the book (there r.colors is recommended).
IMHO pixel values should not be touched.

Cheers

 Markus

-- 
Markus Neteler     <neteler itc it>       http://mpa.itc.it
ITC-irst -  Centro per la Ricerca Scientifica e Tecnologica
MPBA - Predictive Models for Biol. & Environ. Data Analysis
Via Sommarive, 18        -       38050 Povo (Trento), Italy