imited 
n will 
region 
vorld's 
and 
hnical 
| Asia 
"wards 
SCAP 
2000 
Sheet, 
st '93, 
s and 
elated 
d the 
y and 
n the 
Space 
d the 
ofthe 
and 
ecting 
egion. 
1992. 
ce and 
emote 
nental 
n on 
ptimal 
in 
ı and 
toring 
emote 
9512. 
nsing 
PRELIMINARY RESULTS OF THE OEEPE SCANNER TEST 
Prof. O. Kólbl 
Ecole Polytechnique Fédérale de Lausanne, Switzerland 
Commission |, Working Group 5 
OEEPE Special Session 
KEY WORDS : Scanner, Resolution, Imagery, Digital, Digital Imagery, Image Noise, Tone Reproduction. 
SUMMARY 
The aims of this study are the analysis of the tone reproduction of photographic scanners and the development of sim- 
ple test procedures. The most important arguments proposed are the image noise, image resolution, quality of colour 
reproduction, sensitivity of the scanner and the visual aspect of the images. A good scanner should show an image 
noise lower than + 0.03 D for a pixel size of 10 um and a resolution up to 10 um. Furthermore, the dynamic density range 
and the fidelity of tone reproduction should be controlled. 
1. SCOPE OF THE STUDY 
Working methods of digital photogrammetry are more and 
more applied in practice. However the aerial camera using 
photographic films will remain for quite some time superior 
to all digital imaging techniques. Consequently scanners 
play a key role for the conversion of photographic images 
in a digital form. However the draw backs of digital cam- 
eras are also present during scanning and it is of great 
importance do control severely the quality of the scan- 
ning process. The main requirements for image quality 
could be summarized in the following way : 
- Geometric quality 
- Fidelity of tone and colour reproduction 
- Image resolution 
Additionally the user wants instruments, which are easy 
to use and they should operate rapidly. 
Photogrammetric manufacturers, but also the printing in- 
dustry have developed number of scanners of remarkable 
quality and for a user it is not always easy to evaluate the 
different products. However in practical use, one realizes 
very quickly that there are considerable differences in 
image quality between the different scanners. A photo- 
graphic image generally has a rather wide contrast range 
and a high resolution. Very often difficulties must be 
faced when converting those high-quality images into a 
digital form. When comparing an image displayed on a 
digital work station and on an analytical plotter with a 
good optical system, one can see that the analytical plot- 
ter allows a better detail recognition, then on a digital work 
station even when the scanning was done with a rather 
small pixel size. High requirements with regard to image 
quality also stem from automatic image correlation, es- 
pecially when treating low-contrast areas. 
2. OEEPE WORKING GROUP 
ON SCANNER TEST 
In order to develop criteria for a systematic analysis of a 
scanner and to gain experiences on different instruments 
OEEPE (European Organization for Experimental Pho- 
togrammetric Research) has created a working group. 
This working group has organized already a first work- 
shop in November 1993 and conceived a comparative 
test for scanner. Furthermore a questionnaire was pre- 
pared, which should allow to collect technical data on the 
different systems. 
57 
For the experimental test a set of test material was com- 
piled: 
1. Grid plate, grid spacing 1 cm, dark lines on glass 
plate, line width 50 um. 
Objective : Analysis of the general precision of the 
scanner. 
2. Kodak Photographic Step Tablet no 2 (21 steps), 
density range ~ 0.05 - 3.05 D (CAT 152 3398). 
Objective : Analysis of the dynamic density range and 
of the noise of the scanner . 
3. Kodak Ektachrome colour table (from Kodak Colour 
Reproduction Guides Q60A) (CAT 815 5822). 
Objective : Analysis of the quality of the colour repro- 
duction and of the noise in the different spectral re- 
gions. 
4. Fresnel pattern (black-and-white). 
Objective : Analysis of irregularities in geometry. 
5. Black-and-white diapositive copy of aerial photograph 
no 101012 taken with camera Wild RC20 30/4 NATA-F 
no. 17027 on Agfa Pan 150 film (density range of the 
diapositive 0.1 - 1.1 D, copied on film Agfa Avitone 
P3p, aerial flight Penthaz of 15.4.1987). 
Objective : Analysis of the reproduction of a low-con- 
trast image, computation of the MTF and of the image 
noise. 
6. Original black-and-white negative of aerial photograph 
no 5266 taken with camera Wild RC20 15/4 UAGA-F 
no. 13129 on Kodak Panatomic-X film (density range 
0.1- 2.0 D, aerial flight Dübendorf of 8.6.1994). 
Objective : Analysis of the reproduction of a high-con- 
trast original negative, computation of the MTF and of 
the image noise. 
7. Colour aerial photograph no 0439 taken with Wild 
RC20 30/4 NATA-F no. 17027 on Kodak MS2448 
colour film. 
(density range 0.2 - 2.5 D, aerial flight Nyon of 
15.8.1994). 
Objective : Optimal reproduction of a colour aerial 
photograph. It would be desirable that the signalized 
points can be properly recognized and easily mea- 
sured on a digital working station. Furthermore, it 
would be important that details in the shadows remain 
recognizable. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B6. Vienna 1996