The International Archives of the Photoerammetrv. Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
For each terrain point the height profile is determined in 
the principal radial direction and this is followed from 
the terrain point throughout the nadiral determining the 
visibility angles. 
IS 
Centro Perspectivo 
GSD = DSM Cell size 
Px.y 
Siai>«=>V Sioi<a=>l 
Figure 5.- Nadiral Profiles Method 
In this method there are two factors that acquire special 
importance that will influence in the efficiency and the quality 
of the rectification. 
• The profile longitude, which can be determined by the 
maximum height of buildings and the maximum terrain 
slope, assuming a maximum occlusion at the end of the 
used area of the photograph (eliminating overlapping 
areas). 
• The profile resolution, which can be considérer as one 
dimensional, with equal or higher resolution that the 
one from the DSM. 
The problem of orthoimage generation in urban areas is no 
doubt very complex to resolve, because of the great amount of 
information that take part in the process (digital images, digital 
surface models, ...), format varieties and storage structures and, 
obviously, because of the specific cases that exist in resolving 
real problems. 
This is especially interesting if it is considered that within the 
process very expensive equipment are involved, which limit the 
number of tests done, for economic reasons. 
Modules of orthorectification are based at the moment in 
compatible data with the software used for the projects, 
throughout the use of their support files that contain: exterior 
orientation parameters, sensor auto-calibration parameters, 
digital surface models; and they generate images in standard 
format: TIFF with associated files of georreferenciation TFW; 
all these directly exploitable by any Cartography/GIS user. 
The systems has been developed within a C++ environment 
(Borland Builder C++ 2006, Microsoft Visual Studio 2006) 
using specific graphic libraries for image handling (Leadtools 
Medical Imaging Suite v.15). In figure 6 an example of the 
program’s interface is shown. 
m 
<£v* 
A - 
,o. 
CM 
■<DÎ 
Figure 6. General interface of the Altais-LRTO application. 
5. RESULTS 
For the development of this program it was essential the 
information obtained from different tests done throughout the 
project. The framework of this project used, in one way, 
synthetic models in which different algorithms were tested over 
an environment of synthetic images, and on the other way, in a 
posterior stage, using real imagery obtained with the DMC 
digital photogrammetric camera. 
The use of the model allowed to have a perfectly controlled 
environment in which different configuration of “flights” could 
be tested, referring to different scales, overlaps, building 
typology, ... with a terrain information that can be obtained 
easily, very fast and with no cost. Figure 7 shows a detail of the 
model used and on the other way, its definitive location to take 
the images in the Laboratory of Industrial Topography and 
Calibration of Jaén University. Figure 8 shows examples of 
results obtained with the ALTAIS-LRTO program from the 
model’s imagery, which procedure of capture, orientation, etc 
can be consulted in Pérez et al. (2008). 
stages getting ready algorithms. 
Tru« Ortho 
Figure 8,- Example of tests done using the model. RTO+ 
algorithm preliminary results 
Obviously, the objective of having a system to generate 
orthoimages is not reached unless it is applicable to real 
imagery that usually is used to generate urban orthoimages. 
Thus, to complete the tests, a real test was planned and executed 
additionally, providing maximum quality information required 
to posterior verification of results derived from it. Real data for 
this definitive program tests have been captured by the company 
HIFSA using the digital photogrammetric camera Z/I DMC