International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 | 
  
Interna 
oe 
  
  
  
  
  
  
  
  
  
    
   
  
  
  
  
  
  
i 
ES Reference Data | L.N.De 
A Changiang : wt ET | algoritl 
! River River | Lake Veget Building | Road | 00 Ne 
i EX, - ation 
| € River 100 0 0 0 0 LAN 
sesetation Lake 0 l00 | 17 0 | optimiz 
building ; | 
Vegetation 0 0 76 12 l 
road Se — | N. K.. 
Building 0 0 7 84 10 229(1) 
Road 0 0 0 3 89 
- Total 100 100 100 100 100 | 
ES = Accuray 100% | 100% | 76% | 84% 89% 
e vows , * Aula E s 
. . . 8 pas rer. ) | 
Fig.4 The classification image using Artificial Immune Bd : 89.8% 
classifier EUR | 
Kappa 0.8725 | 
coefficient | 
  
  
  
  
EH uuu. 
River 
EX lake 
; BH uo 6. CONCLUSION 
Table 2: Confusion Matrix(AIS Method) 
   
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
building : : x eni Mh 
s In this paper, we synthesize the advantages of artificial immune 
road system, and proposed a new remote sensing image classification 
algorithm using Clonal Selection Algorithm which is a basis of 
the immune system. A quantitative comparison between the 
conventional maximum likelihood statistical classifier and our 
algorithm was demonstrated that the maximum likelihood 
statistical classifier is less capable of discriminating Vegetation 
classes than AIS classifier. Experimental results show that the 
a proposed classification algorithm has high classification 
classifier precision. It is a good and efficient classification algorithm and 
'can be applied to remote sensing image classification. 
Reference Data 
River Lake | Veget | Building | Road REFERENCE 
ation 
River 100 0 0 0 0 D. Dasgupta, 1999. Artificial Immune Systems and Their 
| Lake 0 100 T 0 Application. Berlin, Germany: Springer-verlag. 
T Vegetation | 0 0 59 13 ] F. M. Burnet., 1978. Clonal selection and after. In: Bell G I, 
Building 0 0 20 82 15 Perelson A S, Pimbley GH eds. Theoretical Immunology, New 
Road 0 0 0 4 84 York: Marcel Dekker Inc. pp.63-85. 
Total 100 100 100 100 100 F. M. Burnet., 1959. The Clonal Selection Theory of Acquired 
à Immunity. Cambridge University Press. 
4 Accuray 100% 100% | 59% 82% 84% 
it overall 85.094 LE Hunt, D. E. Cooke., 1996. Learning using an artificial 
i accuracy immune system. Journal of Network and Computer Application, 
oes 0.8125 19(2): pp.189-212 
appa 
coefficient 
  
  
  
  
J. H. Carter., 2000. The immune system as a model for pattern 
RE un 3 : ns A recognition and classification. Journal of the American Medical 
Table 1: Confusion Matrix(Maximum Likelihood Method) Informatics Association, T(3y: pp.28-41. 
J. Timmis, M.Neal, and J.Hunt. ,2000. An artificial immune 
system for data analysis. Biosystem, 55(1/3): pp.143-150. 
L. N. De Castro, F.J. Von zuben, 1999. Artificial Immune 
Systems: Part I-Basic Theory and Application, Tech. Rep-RT 
DCA 01/99. Campinas, SP: State Univerity of Campinas. 
L. N. De Castro, F.J. Von Zuben, 2000a. Artificial Immune 
Systems: Part || - A Survey of Application. Technical Report — 
RT DCA 02/00.