3. RESULTS AND ANALYSIS 
Figure 5 illustrates the output from the multi-layer 
classification process. While figure 6 and 7 represents the edge 
detection map for the study area and the candidate parcels as 
buildings respectively. 
  
  
d000 
| Shadow 
Building 
cis 2e 
2 Asphalt 
  
  
Figure 5. Classification results for the proposed band ratios 
edge map after spur 
E 
A. 
Figure 7. Candidate parcels as buildings 
  
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B7, 2012 
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia 
  
   
    
It is quite clear from figure 7 that the algorithm mixed between 
shadow and vegetation areas with buildings. The integration 
part will account for this confusion as vegetation and shadow 
areas are well defined in the multi-layer classification part. A 
quantitative analysis for the building class (extracted) were 
done by comparing the number of buildings in the original 
image by the complete detected number of buildings being 
detected in the classified images either from pixel-based output 
or from the final integrated output. Table 2 summarizes the 
comparison results. 
  
  
  
  
  
Pixel-based 12 9 1 22 
result 
Accuracy 85.7 83.9 8.8 45,8 
% 
Proposed 14 16 11 41 
method 
Accuracy 189 94.1 64.7 85.4 
% 
  
  
  
Table 2. Comparison of results 
The final result after integrating the two approaches is in figure 
  
Figure 8 final classification results 
4. CONCLUSION 
A method using both pixel and object-based classification 
process was introduced to extract six land covers; vegetation, 
water, buildings, shadows, asphalt roads and bare soil. 
WorldView-2 imagery over Ismailia city, Egypt was used. The 
classification accuracy for five classes out of six was 
acceptable, and it was enhanced for buildings’ class. 
Specifically, buildings’ class accuracy was enhanced from 
45.8% to 85.4% using this technique. This technique was 
applied for many other parts of the large scene, Ismailia city, 
and the results showed great potentials of using this method in 
enhancing the percentage of the detected buildings. 
5. ACKNOWLEDGMENTS 
This work was supported in part by research funds from 
TECTERRA Commercialization and Research Centre, the