ciated with Lin- 
  
  
  
  
um High 
| Vegetation 
iter | Vegetation 
| Water 
  
uistic variables 
avant module at 
uzzy labels and 
membership to 
TEP 4. The three 
ne experts’ fuzzy 
with their transi- 
ium to high are 
hreshold values 
in Table 2. 
ed for Linguistic 
  
bles 
  
  
  
  
  
as been used as 
g the transition 
, as itrepresents 
hip to non-mem- 
s also supported 
tion (simulating 
of human eye is 
j that the maxi- 
rique which gen- 
n accuracy also 
ity function is a 
Further, the bell- 
 Zadeh’s x (pie) 
gain a combina- 
n Figure 2. 
ledge and fuzzy 
that the charac- 
'zy labels repre- 
the overlapping 
ry of land cover 
mbership grades 
ent to which the 
  
   
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
SML b e f L 3 
"T LN | p: | 
a | | | 
r4 | ! | 
o | | | 
iu | 1 
2. i ] 
= | i ! 
| i | 
hE 
S P | L i 
c x3 3 < M 
So. 55 -0.30 woe 00:20 "exw ® 0.70 
Norma! lead Difference Vegetation Index 
® (d [5 e 4$ e | 
- | MON | 
& | | 
9 i i 
uj D Li | 
ge |; | 
ü | | 
| 
| | 
® 11 — NC | Ne 
® Xt eil ch x 9 
Q 69 120 189 240 300 360 
HUE CFalee Color Compoe tte) 
e |o be i f J 
a V 
| 
| 
= | 
% | | 
nis | | 
ge | | 
E | 
| 
{ | 
8 : LN i 
So 5 10 EC 2b 2. 2 38 
TONE CRed bend). 
Fuzzy Labels: LOW - abc; MEDIUM - defg; HIGH - htJ; 
Figure 1: Cherectertistics functions of Fuzzy Lebels 
for NDVI,-HDUE and TONE. 
241 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B8. Vienna 1996