dustrial 
-range 
nt. The 
otimum 
jramm. 
mission 
On-line 
Arch. 
A, Vol. 
hkeits- 
2 from 
insing, 
iultiple 
Yology, 
CRITERIONS OF SELECTING OPTO-ELECTRONIC COMPONENTS OF SYSTEMS INTENDED FOR CONTACTLESS 
AND TRUE TO GEOMETRY MEASUREMENT 
Dr. of Engineering Sience Ralf Goldschmidt 
Bachelor of Physics Werner Borchardt 
Certificated Engineer Uwe Richter 
Rheinmetall Jenoptik Optical Met 
FRG 
Comission V 
Head of Products/Clerk Section 
Specialist 
Project Manager Product Section 
rology GmbH 
ABSTRACT 
People developing opto-electronic systems for contactless and true to geometry measurement not in any 
case have the appropriate conditions of being sure, that they selected the most suitable opto-elctro- 
nic components for their task from a great variety of components offered on the market. On the exam- 
ple of CCDs and frame grabbers interesting measuring 
problems - e.g. linearity errors of CCD measure- 
ment of lateral pixel sensitivity and pixel geometry, investigation and verification of true to 
geometry imaging features of frame grabbers as well 
as quality investigations of video pick-up chan- 
nels of frame grabbers - and measuring results gained on electronic components of different producers. 
are presented and discussed. 
Keywords: Accuracy, Camera, Photogrammetry, Sensor, Standards 
INTRODUCTION 
During the conceptional development phase of 
opto-electronic instruments intended for 
solving measuring, testing and automation 
tasks an optimum selection of  opto-electro- 
nic components - for example CCD and frame 
grabbers for such systems - is of decisive 
importance. 
They determine quality and use value . In 
consideration of the variety of  opto-elec- 
tronic components with their partially most 
different features offered on the market and 
the entirely insufficient  comparability of 
the informative material about them for high 
sophisticated image processing tasks it is 
absolutely necessary to measure their  para- 
meters or to test them under future applica- 
tion conditions (spectral range, dynamic 
range, optical and electronic resolution). 
2. LINEARTY ERROR 
CCD components" linearity is investigated by 
taking their characteristic curves (U = f(E:t), U 
- video signal, E:t - irradiation) when they are 
illuminated by a steady source. Irradiation can be 
altered or by changing irradiance or by varying 
integration time. For solving complicated measu- 
ring tasks you have to find a quasi-linear range 
in case of non-linear characteristic curves and to 
define it for the dynamic range permitted later on 
in the measuring system. If the dynamic range 
limited in this way is insufficient correction 
algorithms have to be applied based upon these 
measuring results. 
Within the scope of the development of non- 
contact metric measuring and 2D-3D image 
pick-up systems in the laboratory of Rhein- 
metall Jenoptik Optical Metrology GmbH a lot 
of different sensible opto-electronic units 
were tested and certificated accordingly. 
We would like to demonstrate here a selected 
number of the most interesting measuring 
problems 
The measuring results gained on most dif- 
ferent components of different producers un- 
der equal conditions and measuring methods 
illustrate the topical | importance and bri- 
sance of an approach like that. 
Thus a decision you have to take for a sys- 
tem conception will be effective and pos- 
sible at the earliest possible moment of the 
development. 
OF CCD COMPONENTS 
he basic idea of the incerpretation method of 
quasi-linear range determination consists in the 
fact that a characteristic curve is approximately 
linear when its rise change tends to zero. This 
means that in case the characteristic curve is 
indicated as an analytic expression (polynom) the 
second derivation must disappear. 
The principle of this method is shown on figure 1l. 
Besides giving the quasi-linear range with the aid 
of the second derivation the first one supplies an 
additional information about differentiated sensi- 
tivity in this range.