satellite images. 
ie pixels are not 
coordinations of 
ation process is 
images resulted 
ly higher degree 
als "Referential 
ht absolutweise 
entsprechenden 
B wird in zwei 
rkundungsdaten 
\auigkeitsverlust 
ngs to draw and 
se of previous 
decisions (Hunt, 
t in which the 
not considered 
elated with the 
ixels and their 
0 enhancing the 
ation of image 
“ATION 
is method first 
oordinates (x,)) 
owchart of the 
might be done 
pixel are valid 
tion process has 
rdinates of each 
  
pixel are related to national or even international grid 
coordinations. The pixel data base is used for archiving 
all previously measured and/or expected information 
about each pixel along with its coordinates within the 
image. The most significant factor in the pixel data base 
design is its ability to include multitemporal information 
about pixels and indices to their trends of possible 
change in the future because any analysis process done 
on pixels depends not only on the present measurements 
but also on measurements from the past. Coordinate 
assignation should be in compliance with the resolution 
of the imaging sensor of the satellite system; that means 
an image of 30 m resolution should be based ona 
coordinate grid of 30 m or smaller steps but not larger. 
The referential classification process is carried out in 
two subsequent stages. In the first stage a supervised 
classification algorithm is implemented according to the 
maximum likelihood rule which is maximum likelihood 
rule which is presented and discussed in details in image 
processing literature (Swain, 1978; Richards, 1986). 
In this algorithm, suppose that there is a training data of 
an image is available for each groundcover type. This 
data can be used to estimate a probability p(x/w;) 
distribution for a cover type that describes the chance of 
finding a pixel belonging to a class w; at a specific 
position x. This could expressed by the term p(x/w) . 
| START | 
GET PIXEL 
  
  
  
ASSIGN COORDINATES 
TO PIXEL 
  
  
P(m,n) 
  
CLASSIFY PIXEL 
(Supervised) 
  
P(m,n,i) 
  
  
  
    
    
Pm.n,T) A 
PIXEL DATA 
BASE 
  
  
GET INDESES OF SIMILAR 
/COORDINATES FROM DATABASE 
  
  
Report P(m, n) as 
  
no 
li fn l 
  
rejected or changed 
pixel 
  
  
ADOPT CLASSIFICATION 
  
Store results 
he 
  
yes 
  
RESULT 
  
  
  
  
  
  
  
  
  
  
  
  
Figure 1, Flowchart of the Referential Classification Algorithm. 
333