International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 
  
4.4 Bari canal 
Four radiometric transects (transects nr.12, 13, 14 and nr.15) 
were traced at Bari canal (Figure 11). Transects nr.12, 13 and 
nr.14 showed correlations greater than 0.72 (transect nr.12 
R’=0.86, transect nr.13 R’=0.72, transect nr.14 R’=0.96). 
Figure 12 show results. 
  
Figure 11. Radiometric transects at Bari canal with lagoon 
depths overlayed. 
  
  
  
  
Depth (m) 
Real batymetry 
» Batimotric points extrac 
—— Estimated bathym: 
A0 
5 y = -2E-07x « 3E05x* - 000136 + 
R! 208552 
  
690 
Distance (m) 
Figure 12. Bathymetry estimation at Bari canal (radiometric 
transect nr.12). R°=0.86. 
4,5 Canesa canal 
Two radiometric transects (transects nr.16 and nr.17) were 
traced at Canesa canal (Figure 13). The best correlation was 
found in deeper water (transect nr.16) with an R°=0.91. Figure 
14 show results. 
  
98 
Figure 13. Radiometric transects at Canesa canal with lagoon 
depths overlayed. 
  
0,000 Y es NAG LEES DE o 
oo ‘10000 20. NX 2.000 40.000 50,000 60.000 70 pon 80.000 90,000 100 £0 
  
  
  
2000 
-3000 
£ 
$8 
$ 
o f 
-4 000 
Real batymetty 
5,000 e Batimetnc points extracted (SGA algorithm) 
—Estimated bathyn 
x - 4E-06x* +0.001x° - 0.0779x 
6.000 Rt=091 
  
7,000 
Distance (m) 
Figure 14. Bathymetry estimation at Canesa canal (radiometric 
transect nr.16). R°=0.91. 
5. ALGORITHM COMPARISON AND 
PERFORMANCES 
Because of a high spatial variability of the seabed and the 
presence of suspended sediments in lagoon of Venice, the Jupp 
DOP model couldn't successfully determine the coefficients of 
absorption. Moreover, the DOP model computes a single set of 
coefficients of absorption for the entire radiometric transect. 
On the other hand, the SGA methodology presented computes a 
set of coefficients of absorption for every layer (step) into 
which is divided the water column. This lead to a more precise 
estimation of the algorithm parameters, and finally, produces 
better batimetric mapping. For algorithm comparison, Figures 
13 and 14 presents punctual bathymetry estimation at S. Erasmo 
littoral testing site using the DOP model (Figure 13) and the 
stratified genetic algorithm developed (Figure 14). 
Internationa 
MIC Are IE 
s — 
Tr eM 
TAM Em 
Depth (m) 
—— 
Figure 13 
Jupp's m 
Depth (rm) 
  
Figure 14. S 
algoritl 
The great at 
his environ 
ecosystems 
contains. 
sediments t 
lowering ai 
disappearan 
environmen 
lagoon of V 
so emerge 
monitoring. 
The method 
Jupp's DO 
environmen 
Venice sho 
derived fro 
the comput 
developed. 
As like as 
algorithm 1 
coefficients 
the correct 
generalizati 
This proble 
lagoon ecos