International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004 
  
  
   
     
   
     
   
    
  
     
   
   
   
    
    
    
    
  
  
  
  
     
   
     
   
     
    
     
   
  
  
  
    
    
Taking into account the low values for the registered intensity 
data and the noise inherent to the laser, the obtained lower 
accuracy for the georeferenced laser scanner model was 
expected. On the other hand, as previously mentioned, 
retroreflective targets could not be successfully used for this 
stage, otherwise better results could be obtained. Moreover, no 
other kind of measurements, like distances between 
recognizable natural features, could be reliably used, because 
most of them were represented by corners or edges, which are 
generally not well surveyed by a laser scanner given its beam 
aperture. Again, artificial targets, if successfully acquired, 
could be more useful even for this goal. 
7. CONCLUSIONS 
In this paper the application of two different 3-D recording and 
modeling techniques, namely image-based modeling and laser 
scanning, for cultural heritage documentation and visualization 
has been presented. In image-based modeling, a consumer 
digital camera was used, then the acquired data was processed 
through a commercial modeling software to create a 3D 
textured model. In addition, a laser scanning model was 
produced with the main objective to compare the two models. 
The main room in the Aquila tower in Buonconsiglio castle, 
featuring walls frescoed by precious paintings, was chosen to 
carry out the comparison between those modeling techniques. 
To this aim control points and check points were provided by a 
total station survey, measuring both natural features on the 
walls and topographic retroreflective targets. However, due to 
difficulties in identifiying the latter targets in the laser scanner 
model, only the natural ones were used for the comparison. 
Nevertheless, in order to increase the accuracy on check points 
selection it is advisable to use artificial targets, whenever 
possible: if successfully acquired by the laser sensor they 
should provide better results than with natural features. Given 
our experience, we suggest to test beforhand the response of 
the laser to the targets being used, since some laser scanners 
could show an unexpected behaviour when dealing with retro- 
reflective targets, which leads to significant measurement 
errors. 
Beside the discussion about accuracy obtainable by the two 
different 3D modeling techniques described here, a question 
remains still open: which technique is the best one? Based on 
the experience gained so far on 3D modeling and taking into 
account the differences between the main factors involved in 
such techniques (data acquisition sensors, data resolution, 
nature of data noise, adopted modeling procedures, etc.), it is 
the opinion of the authors that discussed techniques are not 
competitive but rather complementary. If the user needs only a 
semplified model of the object of interest, in short time, then 
image-based technique is the most appropriate one, as it can 
provide the 3D shape of the object through images, taken even 
by a consumer color digital camera. Therefore this provides a 
relatively fast and low cost technique for 3D modeling. 
Conversely, if not only the global object but also a detailed 3D 
shape of some of its parts are of interest, then in this case a 
laser scanner could be used, being capable of direct acquisiton 
of a huge amount of 3D data in shorter time compared to other 
classical survey techniques. 
Acknowledgements 
This work was developed with the project "Application in the 
survey, store and management of environmental and cultural 
resources of GNSS/INS positioning and satellite, aerial, 
terrestrial photographic and laser scanning data, transmitted by 
DARC, GSM/GLOBAL STAR, INTERNET methods", partly 
financed by MURST (Italian Ministry of University and 
Research) in 2002 as project of relevant National interest. 
National coordinator: Giorgio Manzoni, head of the Research 
unit Antonio Vettore. We also acknowledge the Castello del 
Buonconsiglio for unrestricted access to the room during non- 
visiting hours. 
References 
El-Hakim, S.F., 2002. Semi-automatic 3D reconstruction of 
occluded and unmarked surfaces from widely separated views, 
Proc. ISPRS Comm. V Sym., Corfu, Greece, pp. 143-148. 
El-Hakim, S.F., et al, 2003. Visualization of highly textured 
surfaces, Proc. VAST 2003, Brighton, UK, Nov., pp. 231-240. 
Innovmetric Software Inc., 2004. http://www.innovmetric.com 
(accessed April 30). 
Paquet, .E., Peters, 4, 2002. Collaborative... Virtual 
Environments Infrastructure for E-Business, Proc. of Int. Conf. 
On Infrastructure for e-Business, SSGRRw02, January 21-27, 
Riegl Inc., 2004. http://www.riegl.com (accessed April 30). 
Soucy, M., et al, 1996. Sensors and algorithms for the 
construction of digital 3-D colour models of real objects. 
Proceedings of the IEEE International Conference on Image 
Processing, Lausanne, Suisse. Sept., Volume II. pp. 409-412. 
ShapeQuest Inc, 2004. http://www.shapecapture.com 
(accessed April 30) 
    
P Vis 
Keywords: 
ABSTRAC 
The use of : 
mainly due t 
virtual realit 
traditional p 
effort. Latel: 
Initially, the 
heritage bui 
built docum: 
and surveye 
wasted befo 
modeling, di 
Therefore in 
different las: 
these sensor 
collection, v 
Beside imag 
survey, the u 
applications 
recently very 
produce higl 
time and wit 
these point c 
accuracy for 
environment 
In order to c 
3D surveyin 
photogramm 
Research C« 
Italy) and t 
Ottawa, Car 
models of t 
applying difl 
a very actra 
3D modelin; 
simple archi 
a barrel vau 
facade, tall 
polygonal aj 
easily as no 
ceiling. 
Furthermore. 
is considere 
and Europe 
are complete 
lives of the 
mature age.