The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
1186 
Figure 4: Differences between LIDAR points (white) and DEM 
grid points (red) 
Finally a DEM of the entire testsite was calculated and 
compared to the regular 2m LIDAR DTM. Here standard 
deviations a are achieved, which approximately correspond to 
the ground pixel size after filtering the differences which 
exceed 3a (mostly points on buildings or high vegetation). 
4. STEREOPLOTTING ACCURACY 
As in the case of aerial triangulation the accuracy assessment of 
stereoplotting is done by comparing the coordinates of 
independent check points measured in the field with the 
coordinates of the same points observed in the stereo plotter. 
4.1 Small scale flights 
Stereoplotting of small scale flights, with 45cm GSD, is mainly 
used in the generation of the Topographic Database of 
Catalonia at 1:5.000 scale (BT-5M). The accuracy of this 
database, as indicated in the technical specifications, is lm in 
the X and Y components and 1.5m in the H component in 90% 
of the points that can be well identified in stereo mode. The 
check points used are ground control points from the 
aerotriangulation of projects at larger scales but not used in the 
BT-5M aerotriangulation. The table 5 shows the results 
obtained in the comparison: 
RMS X 
(m) 
RMS Y 
(m) 
RMS H 
(m) 
N.Checks 
Analog cameras 
0.48 
0.51 
0.71 
2178 
Digital (DMC) 
0.52 
0.49 
0.82 
101 
Table 5: Stereo plotting accuracy at independent check points 
for 45 cm GSD 
As in the case of the aerial triangulation check points the 
horizontal accuracy is very similar in case of the analogical and 
digital images; however, in those cases, where the models are 
created along track, the worse b/h ratio leads to a small 
degradation of the vertical accuracy. It has to be pointed out, 
that even with that small degradation in height the BT-5M 
accuracy specifications for the project are fulfilled. 
90% X 
(m) 
90% Y 
(m) 
90% height 
(m) 
Analog cameras 
0.80 
0.83 
1.17 
Digital (DMC) 
0.85 
0.81 
1.36 
BT-5M 
specification 
1.00 
1.00 
1.50 
Table 6: Accuracy at 90% points for 45cm GSD 
4.2 Large scale flights 
Stereoplotting of large scale flights, with 7.5cm GSD, is mainly 
used in the generation of cartography of urban areas at 1:1.000 
scale. The accuracy of this database, as indicated in the 
technical specifications is: 20cm in the X and Y components, 
and 25cm in H component in 90% of the points that can be well 
identified in stereo mode. The results obtained for the check 
points are shown in table 7: 
RMS X 
Cm) 
RMS Y 
(m) 
RMS H 
(m) 
N.Checks 
Analog cameras 
0.105 
0.097 
0.062 
27 
Digital (DMC) 
0.087 
0.091 
0.103 
29 
Table 7: Stereo plotting accuracy at independent check points 
for 7.5cm GSD 
As in the smaller scale, the horizontal accuracy is very similar 
in case of the analogical and digital images and also a 
degradation of the vertical accuracy appears in the case of 
digital images. It has to be pointed that even with that small 
degradation in height the accuracy specifications for the project 
are fulfilled (see table 8). Notice that the vertical accuracy is 
only slightly worse than the horizontal accuracy in the digital 
images and even better than the horizontal accuracy in the 
analog image; we assume that this is due to the fact that the 
vertical component is determined from check points, which are 
part of an extended plane surface, represented in the image by 
more than 10 pixels: pavements, roofs, thick walls, etc.). Thus, 
during stereoplotting, the operator determines the height with 
the help of many pixels which implies a higher height accuracy 
potential compared to check points defined by one or just a few 
pixels only. 
90% X 
(m) 
90% Y 
(m) 
90% height 
(m) 
Analog cameras 
0.173 
0.160 
0.102 
Digital (DMC) 
0.143 
0.150 
0.169 
1:100 map 
specification 
0.20 
0.20 
0.25 
Table 8: Accuracy at 90% points for 7.5cm GSD 
5. PHOTOINTERPRETATION 
To analyze the photointerpretation changes derived of the 
stereoplotting process using images taken by the digital camera, 
only the opinions of well trained operators, with more than 5 
years of experience using images taken by analogical cameras, 
has been taken into account. In the following a list of 
advantages and disadvantages is given in descending order of 
priority. 
The first advantage is because the images contain more 
information in the shaded areas, more information can be 
displayed to be digitized. In the large scale flights to produce 
cartography at 1:1000 scale, the main consequence of having 
more information in shaded areas is that it has been possible to 
extend the flight period along the whole year, including winter, 
giving more flexibility to our production workflow. A second 
consequence, also in large scale flights where the digitized data