Supervisor 
New users 
Imt defaults 
ILWIS to Alex 
Image Proc. 
Default image 
Display 
Pre-processing 
Haze correction 
Destriping 
Feature extraction 
Dump numencal values 
Histogram and statistics 
2-dimensionaJ histogram 
Vegetation index 
Norm band 
Normalization 
AnaglyphicaJ display 
Image calculator 
Monadic 
Diadic 
Triadic 
Filtering 
Image enhancement 
Original-Laplace 
2’Original-Median 
Image smoothing 
Gaussian 
Average 
Median 
Gradient 
Gradient 
Roberts 
Sobei 
La dace 
2 neighbours 
4 neighbours 
3 neighbours 
Rank order 
Local maximum 
Local minimum 
Rank 
Special 
Original-Median 
Local contrast 
Classification 
Sampling 
Box classifier 
.Maximum likelihood classifier 
Unsupervised classification 
Confusion matrix 
Utilities 
Make header 
View header 
Screen dump 
Magnifier (provided with the Archimedes) 
Calculator (provided with the Archimedes) 
Help 
Exit 
Figure 5: Menu system of Alexander 
The following illustrates Alexander’s structure: 
¡Alexander (RISC OS application directory) 
¡Boot, ¡Run, IRunimage, ¡Sprites, Templates 
(standard RISC OS) 
Sprites (additional sprites) 
Users (file containing the user IDs and pass 
words) 
Spointer (directory with mouse pointers) 
Help (Help text) 
SpritePals (directory containing sprite palettes 
for display) 
L (library of function) 
P (directory of programs) 
DATA (data directory) 
defaults (default selections) 
INF (directory containg header files) 
RAS (directory containg raster files) 
S (directory containing user created sprites) 
4 Data1\ ‘Data2\ ... (further directories with 
data) 
4. Courseware 
Remote Sensing and GIS are sets of tools and 
methods. Because they are rather extensive 
sets of methods, it is necessary to know which 
specific tool can be applied to a specific 
problem. Furthermore, it is important to know 
when and in which order to apply the meth 
ods. To do this, it is very useful to have 
models for both the processes involved in 
image formation as well as the application. 
Exercises are application and problem orien 
ted and are taken from several disciplines of 
the earth sciences. One set addresses stu 
dents of geology or physical geography, an 
other students from urban and regional plan 
ning and vegetation sciences, while a third set 
is directed towards cartographers, photogram- 
metrists and climatologists. Rather than intro 
ducing image processing functions as isolated 
concepts, the students are presented with the 
description of an application. They are put into 
the posrtion of a~ researcher or planner wbrcrts- 
given an assignment. All the information 
required to complete the exercise is contained 
within a task description, including available 
data 
functic 
steps. 
proce< 
order 
functic 
shoulc 
model 
The V 
studer 
are nc 
menu 
select! 
requirt 
tions 
desigr 
tions: 
Menu- 
The f 
studer 
compi 
user-fr 
chine, 
or a f< 
burder 
have i 
learn \ 
and wi 
while ’ 
hierarc 
such t 
are to 
On-lint 
An on- 
eve ry 
details 
specifii 
activitii 
functio 
multi-ti 
RISC-( 
Friendl 
Error i 
user ai 
68