| 2004 
point 
due to 
n and 
idence 
raphic 
with a 
in the 
rukas, 
nation 
model 
sumed 
terval 
idence 
rder to 
'essive 
| This 
Yormal 
ent W 
which 
yint in 
retical 
jon is 
ditions 
: HI — 
f 9995, 
iis will 
cV, 
V, then 
ita has 
hat all 
—— ply 
object 
r limit 
own as 
rdinate 
rticular 
1 every 
would 
etween 
) Zero. 
nt will 
wards. 
ta, but 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 
  
  
  
  
  
  
  
41 4 
32 | 
33 | ; dx 
3 1 Moe [d | i | ! j | u l —— dx Ier limit 
S gan | md AU T I) dx upper limit 
= ] T4. sls IY ri m rm rn aper m 
© | pow 5 | | r ME T —— ify Lez linil 
Sh en L l aed dw unnar li it 
S Sl n dv upper Irm 
> ii 
Ud - set of data 
Figure 5. Confidence interval for coordinate discrepancies of Water area object 
4. CONCLUSIONS 
The Maximum Likelihood Classifier is most commonly used 
than any parametric classifiers and is well suited for accurate 
classification. It assumes that the input data are of normal 
distribution and independent. The biggest attention should be 
paid to the semi-automatic land cover interpretation. The results 
of supervised classification show that some confusion has been 
detected within classified image. In order to avoid occurrence of 
such confusions and improve the spatial classification, 
following approaches are suggested: 
- usage of multitemporal images to individualize information 
classes that where confused in a single-data image (a biggest 
influence for a agricultural land cover); 
- usage textural information to improve results of classification; 
- usage GIS procedures based on auxiliary data. 
The determination of vector data suitability for topographic 
maps updating consists of mathematical area calculation of 
topographical objects from reference data as well from satellite 
imagery data. The suitability criteria 76,975 indicates, that more 
then half identified topographic features from satellite 
imagery could be used for map update. However, defined 
suitability criterion depends on accuracy of reference database. 
By the investigation of planimeric accuracy there was 
determined the coordinate accuracy of identifiable topographic 
objects from satellite imagery. Regarding to sample data 
mathematical statistical approach with a probability of 9995 and 
calculation of confidence intervals for each of topographic 
feature types should be applied. All distinctive points have to be 
collected in a random manner. Otherwise the application of 
suggested methodology would be inappropriate. Data amount 
compiling equal sets of points in each group of topographic 
objects is necessary for determination of confidence interval for 
vector data obtained from satellite imagery. 
According to the research of usage of Landsat 7 satellite data 
for land cover registration with integration to reference database 
(depending on the required accuracy), only Water 
(Hydrographic) object class satisfies the accuracy requirement. 
450 
REFERENCES 
ERDAS /magine, 1997. Tour Guides V8.3, ERDAS, Inc. 
Atlanta Georgia, 451 p. 
Digital vector data base of Lithuanian Space Imagery Map at 
Scale 1 50 000 (LTDBK 50 000), 2001. Specification, Version 
1.6, 23 p. (in Lithuanian). 
Ebadi, H., Azizmohammadi, M., 2001. Merging SPOT 
panchromatic data with multispectral data (XS) using wavelet 
and comparing with IHS method. The 20th International 
Cartographic Conference, Vol. 4, ICC2001 Beijing, China, pp. 
669-675. 
Ohlhof, T., 2000. Visibility of VMAP2 data in aerial and 
satellite imagery. International Archives of Photogrammetry 
and Remote Sensing, Vol XXXIII, Part B4, Amsterdam, pp. 
764-767. 
Koarai, M., Kadowaki, T., Watanabe, N., Matsuo, K., 2001. 
Interpretation characteristics of IKONOS imagery for mapping 
and coordinate accuracy of IKONOS orthoimagery. The 20th 
International Cartographic Conference, Vol 4, ICC2001 Beijing 
China, pp.712-718. 
Sjovall, A., 2001. Research of updating digital cartographic 
databases by IKONOS digital satellite images. Geodesy and 
Cartography, No 2, Vilnius: Technika, pp. 61-67 (in 
Lithuanian). 
Vainauskas, V., 1997. Quality control of digital mapping. 
Geodesy and Cartography, No 1 (25). Vilnius: Technika, pp. 
71-77 (in Lithuanian). 
Parseliunas, E., 2001. Geoinformational systems: Technology. 
Vilnius: Technique, 229 p (in Lithuanian). 
Tamutis, Z., Zalnierukas, A., Kazakevicius, S., Petroskevicius, 
P., 1996. Geodesy. Vilnius: Science and Encyclopedia, p. 384 
(in Lithuanian). 
Eurimage. Multi-mission satellite data. Products. 
http://www.eurimage.com/Products/landsat.shtml (accessed 
  
December,2002).