International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004 
  
4.2 Key process steps & results 
- Consider LP as a LuU with its land use class, see figure 6. 
- Land use data is generalized by merging LP, which has the 
same target land use class at level 1,2 and 3 of LUCS (class 
driven generalization). See results in figure 7. 8, and 9. 
- Enhancement with stretching band 3,4,5. Resampling from 
30m into 15m corresponds to panchromatic resolution. 
Knowledge-based slicing method to distinguish premature 
rice, water and other types from DN of band 3, 5, 8. 
Sampling with ground truth training site. Results are in figure 
10. 
- Overlay with land use unit at level 3, which is generalized 
from LR group, figure 10 with line in red colour. 
- Applying maximum likelihood classification with band 3, 4, 
5 with have the lowest correlation in correlation matrix. 
Filtering and vectorizing the detail boundary of LuU at land 
use class of level 4 in LUCS. Results are in figure 11. 
  
Figure 6: LP is considered as a LuU in commune 
Bacly/Lynhan/Hanam 
  
  
  
  
  
  
  
  
  
  
Figure 8: LuU is generated at land use level II 
242 
  
  
  
  
  
  
Figure 10: Landsat 7 ETM+ composite colour band 3,5,8 and 
LuU at class I.1.a generalized from LR group 
  
  
  
  
  
Figure 11: LuU at sub-class of class 1.l.a detecting in 
combination with RS. 
5. CONCLUSION 
This paper has quantified the multi-resolution of spatial data in 
Vietnam LA by analysing the function, mandate and 
responsibility of each management level and in the particular of 
relationship of all disciplines inside Vietnam LA system. The 
results of quantification are shown on figure 3,4,5. 
The result reveals that the ability to solve the question of data 
inconsistency and difficulties when exchanging SD between 
management levels and discipline by abstracting and 
generalizing land use information from LR group to MLU 
group, from communal level to national level. This also opens 
questions relating to the real demand of multi-resolution of SD 
at four levels and two groups of disciplines in Vietnam 
situation, as discussed on section 3.4 of this paper. 
The case study, however, shows a potential solution to combine 
the result of data generalization from LR group and the RS 
ability in detecting land use data at all classes as required by 
Interi 
MLU 
limite 
= Th 
ma 
ful 
int 
Ih 
of 
Ho 
sou 
reg 
Refei 
Fouri 
land : 
conce 
Vol.5 
Zhou. 
Multi 
LNCS 
Zhou. 
Data | 
2750, 
Refer 
Dale, 
Syste: 
Oxfor 
Larssi 
Tools 
UK li 
Molet 
data n 
Vietn; 
Land 
Refer 
GDL/ 
for Cz 
Carto; 
GDL/ 
guidel 
Hesse 
FIC.” 
RSC - 
remot 
invent 
Trung 
generz 
Vietn: 
manag 
17/20( 
Trung. 
Vietna 
GeoIn 
and Al