ul 2004 
  
it these 
  
th ccd 
nces to 
15 and 
ogram 
xels in 
ure we 
tios of 
ns that 
1 scan, 
juality. 
for the 
icantly 
ms the 
ge 2.5 
1e best 
s even 
ze out- 
  
LM 
6 (3.17) 
] (3.12) 
3 (4.30) 
3 (4.39) 
19 
epipo- 
tios to 
  
  
  
  
Ln 
  
2 (4.53) 
) (4.72) 
3 (4.04) 
| (4.70) 
957 
> 1000 
ratios 
  
  
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004 
  
5 um 10 jum 
(b) 
  
15 pm 20 um 
  
  
  
c=0641 
  
  
  
(b) 
  
  
  
  
  
i 
200} 
150} c z(452 
100} 
| 
50 
0 d 
0005 1 15 2 15 
Figure 7: Distance to epipolar line in pixels for films (a) apx (b) delta (¢) t-max and (d) t-pan scanned with 204m and (e) 
ccd with 9um. 
3.2 Edge response 
Table 3 gives the results for the Blonski edge response test. 
The width of edges is shown in pixels and again the 20um 
leads even to better results than using the 5pm images. 
Here the ccd is comparable to 20pm scan. The Luxen test 
(see table 4) behaves the same. 
  
onm 10pm 15pm 20pm 
apx 5.743 2.167 as» 1.451 (395 1.117 6.19 
delta | 5.222 | 2.034605» | 1.58662» | 1.115 4) 
t-max | 5.621 | 2.310045 | 1.496 ee |. 1.138 «o» 
t-pan 5.426 | 2.141 as» | 1.421 os» | 1.104 don 
ccd 9um - - - 1-111 
Table 3: Width of edges for Blonski edge response test in 
pixels (values in the brackets are the ratios to 5j/m scan). 
  
  
  
  
  
  
  
  
  
  
oum 10pm 151m 20pm 
apx 0.266 0.424 2.5 0.529 (356) 0.601 (5.10) 
delta | 0.253 | 0.418 em | 0.530 «x» | 0.595 55» 
t-max 0.261 0.418 (256 0.518 695 0.580 (4.94) 
t-pan 0.247 | 0.414 as | 0.536 am | 0.613 «io 
ccd 9pm - - - 0.623 
  
  
  
  
  
  
  
Table 4: Width of edge magnitude ellipse from Luxen test 
normalized to the interval [0, 1] (values in the brackets are 
the squared ratios to 5jum scan). 
3.3 Siemens star test 
In this test we used a Siemens star test target with 72 wed- 
ges. The minimal possible radius according to the Kell fac- 
tor is given with min = E 2 = 16. Again the 20pm 
scan gives same results as the 5j/m one and ccd performs 
equivalent. 
  
3.4 Noise estimation via entropy 
Noise is measured using two homogenous image patches. 
The one within a bright area (table 6) and the other one 
within a dark area (table 7). In this test the ccd image is 
1139 
  
Sum | 10pm | pum | 20pm 
apx 64 33 ao 22 on 16 (400) 
delta 66 32 206 | 19 64m | 16 ann 
t-max 65 3560s | 21610 | 16 os 
t-pan 63 36455 | 22036 | 18 050 
ccd Yum - - - 16 
  
  
  
  
  
  
  
  
Table 5: Minimal resolving radius for the Siemens star test 
in pixels (values in the brackets are the ratios to 5/4 scan). 
radiometrically scaled to 16 Bit to make the results com- 
parable. Noise in the digital sensed image is significantly 
lower in comparison to the analog one and this is also the 
main reason why the geometrically test yields better results 
for the ccd images. 
  
oum 10pm | 15um | 20pm 
apx | 1155111 53911 332 | 11.160 
delta | 11.968 | 11.962 | 11.869 | 11.778 
t-max | 12.201 | 12.166 | 12.052, | 11.954 
t-pan | 12.462 | 12.415 | 12.310 | 12.217 
CCD 9pm - - - 7.608 
Table 6: Entropy within a bright homogenous image patch 
in Bits. 
  
  
  
  
  
  
  
  
  
oum lOum | 15pm | 20pm 
apx 12916 | 12.203 1 11422 | 10.730 
delta | 12.797 | 12.185 | 11.426 | 10.765 
t-max | 12.709 | 12.049 | 11.305 | 10.709 
tpan |.11.773 | 11.360 | 10.832 | 10.337 
CCD 9,um = fi = - 8.599 
Table 7: Entropy within a dark homogenous image patch 
in Bits. 
  
  
  
  
  
  
  
  
4 CONCLUSION 
We have compared film-based images scanned with 5, 10, 
15 and 20pm and ccd-based images. In general all four 
used films perform quite similar whereas the Kodak T-PAN 
and Kodak T-MAX are a little bit better in comparison to