n is not 
ormation 
squency. 
can be 
ITF). 
45 
is PC is 
| of the 
ns, that 
transfer 
] during 
the PC 
store all 
nade by 
vriter is 
D. 
ig takes 
der the 
it with a 
line by 
used by 
e photo 
has not 
en also 
pes and 
scanned 
correct 
loss of 
made it 
year of 
2d from 
may be 
ne area 
oard to 
ally the 
ns if the 
An investigation of the warm-up of the lamp and the 
sensor has shown a strong influence during the first 3 
minutes. The gray values of an area of the gray scale has 
changed from the switch on to 3 minutes later from 147 to 
179. After 3 minutes stable conditions have been reached 
with no stronger change than 2 to 3 gray values. The 
remaining changes may be caused by a change of the 
voltage. 
7. Conclusion 
The CCD-lines and arrays do have a limited dynamic 
range in relation to aerial images, by this reason it is 
necessary to use the optimal settings for the scanner. The 
time consuming empirical determination can be replaced 
by a pre-scan with a desktop scanner. Based on the pre- 
scan with a developed table the settings can be found 
within one minute. Nevertheless a loss of information will 
happen in the dark parts. By this reason problems are 
existing in scanning photo negatives. With photo positives 
and color images which shall be scanned only black and 
white, no real radiometric problems are existing. 
The geometric accuracy of the RM1 corresponds to an 
analytical plotter, that means, it is sufficient also for jobs 
where precision is required like image correlation and 
block triangulation. 
The pixel size of 12um is corresponding to the usual 
resolution of aerial photos. A smaller pixel size will not 
improve the quality of the digital image. Because of the 
large file size of the digitized images (~360 Mb for one 
channel, 1.06 Gb for the 3 color channels together) very 
often the pixel size is reduced to 24pm. 
References 
Jacobsen, K.1980: Attempt at Obtaining the Best 
Possible Accuracy in Bundle Block Adjustment, 
ISP Hamburg 1980 und Photogrametria 1982, 
p 219 - 235 
Jacobsen, K. 1982: Selection of Additional Parameters by 
Program, ISPRS Com lll, Helsinki 
Jacobsen, K 1984: Analysis of Remaining Systematic 
Image Errors, ISPRS Rio de Janeiro 
Bósemann, W., Jacobsen, K. 1995: Kalibrierung digitaler 
Bildaufnahmesysteme, DGPF Jahrestagung 
Hannover 
Wehrmann, H. 1995, Kalibrierung des Photoscanners 
RM1 Rastermaster, Diploma Thesis, University 
of Hannover 
77 
RM 1 
Ne 
WEMIILI & ABBOCIATES 
Figure 12: Rastermaster RM1 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B1. Vienna 1996