Table 2. Percentage of points for various classes of absolute differences between manual and automatic measurements (in m). 
  
  
  
  
  
  
  
  
  
  
Version 0-2 2-3 3-4 4-6 6-9 9-15 >19 
DPW 770 (whole image) 85.65/ 5.35/ 2.80/ 2.69/ 1.90/ 0.94/ 0.68/ 
all points / points with code > 32 90.12 5.57 2.20 1.55 0.47 0.09 0.00 
DPW 770 96.75 1.73 0.63 0.40 0.28 0.18 0.05 
(region without big blunders) 
VirtuoZo (whole image) 89.17 3.99 2.26 1.87 1.05 0.81 0.85 
VirtuoZo 97.70 1.38 0.35 0.25 0.18 0.15 
(region without big blunders) 
  
  
  
  
  
  
  
-17m 0m 10 m 
Fig. 4. Accuracy analysis in the region without big blunders. The differences between manual results and DPW 770 (VirtuoZo 
results were very similar) are displayed as grey values. The range -4 to 4 m has been strongly stretched for better 
visualisation. 1 ...big errors due to steep slopes, 2 ...errors at the tip of the glacier tongue. 
  
Fig. 5. Low quality code of DPW matching results: Points with quality code < 33, i.e. unreliable points, are displayed in white and 
overlayed on an orthoimage. They mostly lie in areas with shadows or no texture (compare to Figure 3). 
processing phase. It will also compensate for y-parallax, perform 
epipolar resampling on the fly, and use more than two images. 
4.2. DTM Editing 
Efficient, fast and comfortable methods for the automatic 
detection and exclusion of blunders are still missing. As an 
example big errors (so called spikes) could be easily detected and 
108 
corrected by a median filtering but such functions are not 
provided or not correctly implemented by these two systems. 
Since some blunders may always remain in the data set, a manual 
editing will still be necessary. The tools provided by digital 
photogrammetric systems must be such that all blunders are 
included in a set of suspicious points that is presented to the user, 
permitting him an easy manual or semiautomatic 
remeasurement. Otherwise, the user must visually control 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996 
  
  
  
hur 
ind: 
DP 
cod 
inte 
the: 
ana 
ca. 
res 
sevi 
cou 
int 
are 
Vir 
dis] 
inc] 
inc. 
coc 
me: 
and 
led 
blu 
DT 
Bot 
vie 
and 
blu 
wh 
bef 
etc. 
inr 
ceri 
spe 
ver 
bre 
Coi 
she 
phc 
the 
lik 
rec 
mo 
ime 
dat 
the 
tri 
dig 
ort] 
ort] 
De: 
foll