IX-B3, 2012 
search Program of 
tion Program of 
mmission (No. 
subset of the data 
carried out by the 
] Remote Sensing 
r. And we would 
on Urban Object 
n, which provided 
aillard, S. Benitez, 
c filtering and 3D 
. Commission III 
ROM. 
ht Adaptation and | 
egmentation. [EEE 
-1123 
ital aerial camera 
Photogrammetrie 
marker controlled 
SM and imagery. 
), 1441-1452. 
utomatic building 
afer algorithm. 
ensing 72 (4), 395- 
nplexity measure : 
es of IEEE Annual 
mation Processing 
ntation and object- 
building class from 
3(8): 1076-1087. 
Tracing boundary 
/ision Computing, 
ing the Dempster- 
‘and multi-spectral 
usion 6(4),283-300. 
. A comparison of 
ion from airborne 
Topics in Applied 
,11-20. 
:m house detection 
shape, and context. 
| of remote sensing 
f Photogrammetry 
xcitatory globally 
sactionson Neural 
  
  
wang, D.L., Terman, D.,1997. Image segmentation based on 
oscillatory correlation. Neural Computation 9, 805-836. 
Wang, D.L. 2007 .Computational scene analysis. In Challenges 
for Computational Intelligence, W. Duch and J. Mandziuk, Ed, 
DSM raster data 
Springer, Berlin. 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B3, 2012 
XXII ISPRS Congress, 25 August - 01 September 2012, Melbourne, Australia 
Wang, Z. Schenk T., 2000. Building extraction and 
reconstruction from LiDAR data, International Archives of 
Photogrammetry and Remote Sensing, Vol. 33, Part B3. 
Amsterdam, The Netherlands, unpaginated CD-ROM. 
  
     
DSM Height textures based on 
  
  
Initialization 
Calculation of coupling weights 
Leader cell selection in Building targets 
Determination of Global inhibitory 
    
   
  
  
  
LEGION for DSM 
Segmentation 
the GLCM 
extract buildi 
DSM Roughness to determine 
S Global Inhibitory value 
  
  
  
  
r 
  
    
Self-excitable cell detected 
  
Self-excitation 
  
segmentation of one region 
  
—— 
  
  
YES 
  
   
   
    
  
Excitable 
Cell detected 
  
Excitation of 
NO | dependent cells 
Inhibition to | 
Region Labeling 
  
  
  
  
  
  
  
  
  
  
  
  
  
Figure. 1 Flow chart of Extended LEGION Segment for Building 
  
  
( DSM Segmentation and Building Extraction ) 
  
Extraction from DSM 
    
Suburban 
Figure 2. DSM of experimental data 
  
A2 
Suburban 
Figure 3 GLCM homogeneity result of experiment data(homogeneity in light)