| International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 
  
| 
| 2. CHANGE DETECTION ANALYSIS 
Using 1 meter spatial resolution IKONOS imagery, in detection 
of collapsed buildings after the earthquake, is extremely 
valuable in terms of time and man power. 
| Detection of collapsed buildings after the earthquake can be 
| easily done by satellite images acquired on different dates. One 
of the methods to determine this change is, visual interpretation 
of buildings one by one and the other is change detection 
analyses method. 
   
2.1 Visual Interpretation 
  
AFTER 
First of all, in visual interpretation method, to detect collapsed 
building, satellite images acquired on different dates are Figure 3. Digitization of large areas. SULTANDAGI 
Is compared and the buildings that have damaged or collapsed are Earthquake, CAY Center and CAY Industrial Arca ird 
s digitized. It could be possible to make digitizing in higher FEMMES 2002 Colla sed buildings ac detected i al 
ie scales as in building bases or for large areas. The detection of Yeu ap figs are deiecta Dy visua 
le collapsed buildings by visual interpretation and the digitized interpretation: 
K buildings can be seen in Figure 2 and 3. 
a 
Or 
be 2.0 CHANGE DETECTION ANALYSIS METHOD 
Other change detection method is; change detection analysis 
method. Satellite images acquired on different dates are 
required for this analysis. To increase the efficiency and success 
of the method, it is wiser to perform masking. For masking, 
vector data of the buildings in the study area is needed. By this 
method, change detection analysis would enclose only the 
related areas (building boundaries), where the mask is present 
and the results would yield the change only in these areas. The 
flow chart showing the steps of the change detection analyses is 
BEFORE AFTER shown in Figure 4. 
  
  
Figure 2. Digitization in building scale. SULTANDAGI 
| Earthquake, CAY Center and CAY Industrial Area 3rd 
| February 2002. Collapsed buildings are detected by visual 
interpretation. : Satellite Satellite 
E Imagery Imagery 
B" ah. 
  
  
  
  
  
  
  
  
| As seen in Figure2, buildings could easily be digitized by using 
| high resolution imagery. After that digitization process the Mask PA 
| produced vector layer could be correspond with a new image Laver 
and collapsed buildings could be detected visually. 
  
  
  
  
  
  
  
  
Change 4 
Detectio 
  
  
An other digitization could be done on larger areas like 
industrial zones as shown in Figure 3. This process will prevent 
time consuming for digitization. 
  
  
  
Polygon 4 
  
  
  
  
Figure 4. Flow chart of change detection analysis 
mite: 
  
BEFORE 
1269