008
mM
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
Area 3: some hilltops were developed.
FSmk
Figure 5. Old aerial photo (1/20,132 13 Mar 1956)
Figure 8. PRISM Ortho Image (31 Jan 2007)
3.2 Analysis
In the present study, a numerical classification was conducted
to grasp visual information of typical topographic changes in
different area. The quantification analysis estimated the below
characteristics: basic landform, location and land use and
topographic change.
Figure 6. PRISM Ortho Image (31 Jan. 2007)
Area 4: A dam was constructed.
Figure 7. Old aerial photo (1/39,610 15 Feb 1948)
Question
Intensity of Validity
a N
1
2
3
4
5
Is area covered with flat
land?
No
Yes
Indented?
No
Yes
Elevated?
No
Yes
What is the scale of
change? (km 2 )
Less
than
0.1
1
10
Larg
er
than
What is the height scale
of change?(m)
Less
than
5
10
15
Larg
er
than
Table 1. Sample of threshold
The analysis of the landform change showed that changes
caused by human activities were extending along the roads or
river basin. The major changes caused by human activities (e.g.
construction of new towns etc.) were extended to gradual slopes
and hilltops and the terrain changes influenced significantly for
geo-rectification. Furthermore, there are some acute changes on
steep slopes in mountain regions and were caused by disasters
or disaster prevention measures. Especially for all widespread
change, it is difficult to recreate the old terrain model from
PRISM DTMs automatically, whereas some references about
the change were substantially available.
4. ASSISTING FOR THE COLLECTION OF GCPS
Based on the trend analysis, we tried a streamlining for the
collection of GCPs. Most common shape of existing features
after 60 years is linear shape. And point of intersection of lines
is easier to pinpoint the pixel. For this reason, we selected some
common points from old and new linear features. Those points
421
iff-.
