The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Voi. XXXVII. Part B3a. Beijing 2008 
DTM network,DSM network and NDSM 
network are shown in figure 1,2,3 respectively. 
jn****** 
' 
t 
w 
if DTM 
I I 1140.361 - 1140.834 
I I 1140.834 • 1141.307 
I 1141.307 - 1141.78 
f ! 1141.78- 1142.253 
1142.253 - 1142.726 
■ 1142.726 - 1143.199 
1143.199 - 1143.672 
1143.672 - 1144.145 
m 1144.146 - 1144.618 
■Ü No Data 
m " ^ 
if' nDSM 
I I -0.215- 1.413 
I I 1.413- 3.04 
I I 3.04- 4.667 
¡Ü 4.667 - 6.295 
HU 6.295 • 7.922 
B 7.922- 9.55 
9.55 - 11.177 
11.177 - 12.805 
8 12.805 - 14.432 
N o D ata 
Figure 1. is two dimensional image of DTM network 
Figure 3. is two dimensional image of NDSM network 
3. Morphological Reconstruction 
Mathematical morphology suggested by 
Matheron and serra,first in 1970,this method is 
used to analysis images and signals and 
expanded according to series theory. 
Also it has developed in digital image 
processing as image enhancement ,edge 
detection,context analysis and noise removal in 
images.in addition, it can be usefull for 
extraction of peaks in digital models such as 
DSM, DTM and NDSM. 
one of the efficient operator in mathematical 
morphology is reconstruction that is base of 
algorithm to finding of peaks in NDSM,so we 
will explain it’s conception. 
3.1. Surface reconstruction for obtain the 
demes network 
one of the efficient application of surface 
reconstruction is providing of demes 
network.infact for performing a reconstruction 
surface,we need mask and marker layer that 
marker layer is lower to mask layer, 
in this application , we assume specified layer 
such as NDSM as well as mask layer and then