iphical 
use all 
ftware 
erface 
ss to 
litional 
image 
at the 
SCRE 
Ictions 
| from 
e files 
»tailed 
| now. 
ns the 
ent of 
of the 
loads 
lors or 
mages 
ement 
device 
1S are 
mpare 
can be 
time. 
gitized 
with 
lowing 
h the 
tewise 
traight 
it any 
mation 
| input 
image 
image 
ıtrol of 
INEPL 
es are 
o add 
d and 
igitizer 
nerate 
for the 
). With 
ich will 
be matched in the calculation of image mosaics. The 
database exchange file can be converted with CLIPPER 
database software to dBase standard format. 
This interface is important for the data integration into 
GIS. With this link object and attribute collections can be 
included into ARCVIEW (Buhmann, Schaller, Bachhuber, 
1996; Wiesel 1996). 
4. EXAMPLES 
Most of the following image examples are used in 
education and training at IPl, Hannover. 
To reach better throughput they are limited to image 
dimensions up to 1024x1024 pix. This is only to get fast 
results while the education is going on. The programmes 
normally have no limitation in the image dimensions. 
4.1 Digitized aerial photographs 
To show the large amount of details in a vertical aerial 
photograph, an image of the Great Royal Garden in 
Hannover has been digitized with a resolution of 150 dpi 
(Fig.1). 
  
Fig.1, Digitized aerial photograph 
To get all details it is recommended to scan with 600 up 
to 1200 dpi, but for educational purposes it is a good 
141 
compromise to reduce the required amount of disc space 
by using a reduced resolution. 
In comparison to this image an oblique amateur 
photograph has been digitized. 
This is used to demonstrate the effect of perspective 
transformation in the image to image registration (Fig.2). 
  
Fig.2, Oblique photograph 
4.2 Digitized maps 
For quality control purposes a digitized city map in the 
scale 1 : 20.000 is used (Fig.3). 
  
Fig.3, Digitized city map 
For the registration of satellite images a topographic map 
in the scale 1 : 100.000 has been digitized. 
With the program ONSCRE several objects in vector 
presentation are measured in the digital image. This 
vector information can be superimposed with the rectified 
aerial image. This is a fast approach for a first visual 
quality control. 
For the final quality control tiepoints in image and map 
have to be measured and residuals can be calculated and 
displayed with additional software. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B6. Vienna 1996