Figure 1 
Test site of Section 2 (Map of 1987) 
  
Figure 2 Test site of Section 2 (Orthophoto of 1994) 
were measured in this measurement and this DEM was used to 
check the accuracy of automatic matching. Namely following 
four kinds of DEM were prepared, the posting interval of which 
is 2.5 m on the ground for every DEM. 
(1) DEM made from old photos (1987) by stereo matching 
(2) DEM made from new photos (1994) by stereo matching 
(3) DEM made from contour lines of 1:2,500 map (1987) 
(4) DEM made from old photos manually using a stereo-plotter 
Five test area were chosen in the test site. Each test area is 150 
548 
m X 150 m in size and therefore has 61 X 61 grid elevation 
points. The land use and its change of test areas are shown in 
Table 1. 
Statistical parameters of height in each test area were calculated 
and examined whether these can be used as the indication of 
land use change. Statistical parameters were also calculated for 
changed test area size by dividing the areas into four or nine in 
order to find out appropriate area size for this analysis. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996 
T 
22. 
22. 
DE 
the 
show 
maté 
for t 
  
wher 
It ca 
heigl 
indic 
withi 
consi 
accur 
areas