ltural 
+. Lhe 
SLAR 
atisties, 
ment, 
tudy 
ata 
pected 
1d be 
ean 
faces 
f 
DFVLR, 
1980. 
Space 
seful 
and 
India, 
rpre- 
y y 
„412 
angle 
Ped), , 
ering 
^amme - 
.DFV LR 
nd 
| at 
rent for 
tal 
ich 
RTE 
INTERPRETABILITY OF LINEAR PHENOMENA ON SEASAT-1 
IMAGERY IN THE WESTERN COASTAL ZONE OF BELGIUM 
IN RELATION TO THE AZIMUTHAL AND RANGE RESOLUTIONS. 
H . DUFOURMONT 
(I.4,0.N.L) 
Laboratory for Regional Geography and Landscape 
Studies 
State University of Ghent, Belgium 
ABSTRACTS 
In this study, the visual interpretability of linear features on SEASAT image- 
ry was checked by means of comparison of variables two by two. The orientation 
of the flightline seems to influence the detectability of lines on the image. 
This assessment harms strongly the visual interpretability of linear features. 
This paper deals with the visual interpretation of SAR imagery 
taken by the SEASAT-1 satellite. The images are digitally processed on the 
MDA digital Seasat S/W-processor at DFVLR (Oberpfaffenhofen, Germany). 
The available negatives have a 1/250.000 scale, 4 looks and a 25 m resolution 
for both range and azimuth directions. 
The testzones chosen for this study are situated in de Western 
coastal zone of Belgium, between Koksijde and Ostend. In this testzone two 
geographical regions are present : 
1. A narrow coastal strip, comprising a Sandy beach, bordered 
by a sanddune-belt that changes landwards into a lower dune 
landscape. This strip is built up, quite densely, by touristi- 
cal accomodations. 
2. A "polderland"-area with a typical open field landscape, 
topographically very flat, about sea level and characterized 
by an agricultural occupation on moist clay soils. 
A preliminar visual analysis of the image showes a large number 
of linear features, most of which can be identified as human constructions. 
Visual interpretation has been chosen because it is the first approach to the 
available SAR images. 
Considering the direction of the linear features, a striking 
difference in density can be observed according to the groundtrack of SEASAT. 
Most linear features show parallelismto the azimuth direction, whereas the 
range and diagonal directions show a restricted number of linear features. 
In order to evaluate the impact of the difference in resolution 
on the interpretability of linear phenomena, we have elaborated several tests. 
Following variables were taken into consideration : 
1. The orientation of the linear features 
2. The lenght of the linear features 
, 
. 
, 
249 
M ne S nr aa i RR RR I 
bl uiu EE OE SE HRS SE DL 0 ER