No subject

[Net.Noise owner]

Newsgroups: info.grass.user
Path: zorro.cecer.army.mil!shapiro
From: shapiro at zorro.cecer.army.mil (Michael Shapiro)
Subject: i.maxlik
Message-ID: <C9tJF9.1oM at news.cecer.army.mil>
Keywords: maxlik reject
Sender: news at news.cecer.army.mil (Net.Noise owner)
Organization: US Army Corps of Engineers Construction Engineering Research Labs
Date: Wed, 7 Jul 1993 23:47:33 GMT
Lines: 22

The "reject" map for i.maxlik is a chi-square measure of how well
each pixel fits into the class to which it is assigned. I am not
a statistician so I can't tell you more than that. However I
can say that the "reject" map has been used for single class
classifications. John Isaascon (isaacson at zorro.cecer.army.mil)
has used this feature to identify archeologic sites.
He used i.class to extract the signature for one site he
could see in an image. Then he ran i.maxlik on this single-class
signature. Of course the output map had all pixels assigned to
this class. But then he used the reject map, selectively setting
the MASK to a subset of the categories in the reject map and then
viewing the classification map and found that this technique
enabled him to detect other sites in the image.

He did point out to me that the labels for the categories in the reject
map seemed backwards. If you test this use of the "reject" map, you
might want to keep this in mind (and let me know what you find out).
-- 
Michael Shapiro               shapiro at zorro.cecer.army.mil
U.S. Army CERL                (217) 373-7277
P.O. Box 9005 
Champaign, Ill. 61826-9005