Structural Modeling of a Historic Castle Using Close Range

Photogrammetry
VINCENZO BARRILE (*), GIULIANA BILOTTA (**), ENZO D’AMORE (*),
GIUSEPPE M. MEDURI (*), SANDRO TROVATO (*)
* DICEAM - Faculty of Engineering
“Mediterranea” University of Reggio Calabria
Via Graziella Feo di Vito 89100 Reggio Calabria, Tel +39 0965 875301
** Ph.D. NT&ITA – Dept. of Planning - University IUAV of Venice
Santa Croce 191, Tolentini 30135 Venice
ITALY
[email protected], [email protected], [email protected],
[email protected], [email protected]

Abstract: This paper presents a methodology for building surveying and structural modeling of buildings of
monumental interest. The complex geometry of these structures requires appropriate techniques to support the
diagnosis, health monitoring and structural analysis phases for the validation of the expected performance under
future earthquakes. This study aims to investigate a low-cost technology to generate accurate 3D structural model
of large historical buildings using photos taken by handheld digital cameras and commercial software. In this
regard, a multidisciplinary approach, that integrates close-range digital photogrammetry and finite element
analysis (FEM), is tested to the case study of a XV-XVI century masonry castle in southern Italy (Marina di
Gioiosa Jonica, Reggio Calabria). Furthermore, to assess the accuracy of the modeling process the geometry of
the generated model, we compared it against classical buildings surveying techniques. Finally, a linear FEM
analysis is undertaken using the 3D model to show the potential of the adopted procedure for the purpose of
structural analysis of a complex structure under earthquake loads.

Key-Words: Masonry structures – Building surveying – Cultural heritage – Photogrammetric techniques –

Geometrical survey – Structural modeling – Earthquake analysis – Historical buildings

1 Introduction study of Torre Galea, shown in Fig. 1, a relevant

Cultural heritage requires the implementation of Italian XV-XVI century masonry castle built in
suitable survey techniques for preservation and Marina di Gioiosa Ionica, Reggio Calabria. The
promotion purposes ([1][2][3]). proposed procedure aims at the implementation of
The introduction of new measuring devices such the already mentioned digital photogrammetric
as 3D laser scanners, spherical photogrammetry, techniques for the semiautomated generation of an
structure-from-motion photogrammetry and the accurate 3D model. The generated model is then
latest methods of image-based modeling produced a converted into a finite element model for the analysis
strong change in the acquisition, treatment and of building performance under static and dynamic
restitution of metric information. These new loads and for structural health monitoring purposes
techniques allow for the construction of digital ([13], [14], [19], [17]). Furthermore, to assess the
photo-realistic 3D models of complex buildings and accuracy of the modeling process the geometry of the
structures hardly suitable for traditional surveying generated model is compared against classical
techniques ([4],[18]). Models so obtained can be buildings surveying techniques.
integrated in a novel information system capable to The first phase of the proposed procedure, digital
manage complex and typologically heterogeneous photos of the castle, taken by a handheld digital
data relevant for the monitoring and preservation of camera, are processed by a stand-alone software
cultural heritage ([5],[6]). To this purpose, in this (Photoscan by Agisoft [8]) that performs
paper, promising close-range digital photogrammetric processing of digital images and
photogrammetric techniques are applied to the case generates 3D spatial data. Then, for eliminating mesh
defects, the generated 3D model is imported into
Geomagic Studio software [9]. This software allows Digital Image 3D model
for the editing of the point cloud and the mesh. Acquisition construction
Geomagic also provides advanced surfacing editing
functions (e.g. decimation), in addition to accurate
functions of 3D data processing.
Finally, we imported the amended model into a Construction of
Polygon mesh
the model with 3D
FEM software (SAP2000 V.17 [11]). Linear static improvement
surface objects
and dynamic analyses are performed to investigate
the behavior of the structure under gravitational and
earthquake loads.
Finite element
model for
structural analysis

Fig.2. Algorithm of the adopted procedure for the

generation of a 3D structural model from a close-
range digital photogrammetric survey.

2 Construction of 3D Model
The procedure of photographs processing and 3D
model construction comprises four main steps as
described by Photoscan developers.
1. The First step is the camera alignment phase.
At this step, PhotoScan seeks common points on the
photographs to be merged. The result of this step is
a cloud of sparse points (Fig.3) and a series of camera
positions. The point clouds represents the alignment
results between photos and will not be used directly
in a further procedure of construction of the 3D
model. However, it can be exported for further use in
external programs. The construction of the 3D model
by PhotoScan is based, instead, on the set of data
related to camera positions.
2. The next phase is the construction of dense
point cloud (Fig.4). Based on the estimated positions
of the camera, PhotoScan generates a point cloud
more dense and detailed. This point cloud can be
modified and classified before being exported with
the generation of a three-dimensional mesh model.
3. Then we proceed with the construction of the
mesh (Fig.5). PhotoScan reconstructs the surface of a
Fig.1. Views of Torre Galea- Marina di Gioiosa
3D polygon mesh representing the object based on
Jonica (RC).
the dense point clouds obtained from the previous
step (Fig.6). Generally, there are two algorithmic
methods available in PhotoScan that can be applied
The flow chart of Fig.2 shows the algorithm that,
for the generation of 3D meshes: Field Height - for
starting from the digital images, yields to the 3D
planar surfaces, or Arbitrary - for any object type.
structural model.
4. After building the polygonal network, it may be
necessary to make some adjustments. PhotoScan is
able to make corrections, such as decimation of the
mesh, the removal of isolated components, the
closing of holes, etc. When a more complex and
detailed editing is pursued, an external editing
software can be used. In this regard, PhotoScan
allows the export of the mesh for further editing with
other software. Then, the refined model can be
inported in PhotoScan through the most common
interchange formats.
5. After the geometry and the mesh has been
reconstructed, it can be textured (Fig.7) and/or used
for the production of orthophotos. PhotoScan
implements several texturing algorithms, described
in detail by the software developers.

Fig.5. Mesh.

Fig.3. Point Clouds. Fig.6. Texture with the position of the camera (blue
square).

Fig.4. Dense Point Clouds. Fig.7. 3D model with texture.

 We now scaled the obtained 3D model. For the The steps to follow in order to improve the mesh
scaling process to be successful, a precise measuring are:
base has to be determined. In our case study, the span 1. Import the model (STL) within Geomagic
of the castle gate was selected as a reference basis. Studio and set the unit of measurement.
In order to scale the model we define two markers 2. The software automatically recognizes the
(Fig.8) that allow for the measurement of the distance presence of mesh and asks if you want to launch an
between two known points. Then we proceed to analysis mesh doctor. The mesh affected by errors are
create a "scale-bar" and to change the known identified directly on the model with red areas
distance. (Fig.9).
3. If necessary, we can rescale the model using the
specific tool available in Geomagic (Fig.10).
4. The last step is exporting in dxf format for CAD
editing. From here, we create a file that we imported
into Autocad, from which we have to create a new
layer and transfer the imported model on it. Our goal
is to build 3Dface objects that are geometric elements
that can be imported into SAP2000 or other FEM
software. Geomagic software automatically exports
3Dface, the only problem is that it automatically
creates a layer named "layer 0". Sap2000
unfortunately requires the use of a different layer that
can be created using a CAD software.

Fig.8. Markers with flags.

From the 3D scaled model, we want to move to

the FEM software for the structural analysis. This
step is not directly possible because the exported
mesh is affected by errors related to discrepancy
between the various points constituting the mesh; for
this reason we cannot import directly into a software
FEM the 3d model. We must make an intermediate
step to eliminate the defects of the mesh.
We can export the 3D model in different
interchange formats. In this work, we used the STL
format (Stereo Lithography interface format or
Standard Triangulation Language)

3. Polygon Mesh improvement

For eliminating defects of mesh, we proceed
exporting the 3D model from PhotoScan in STL
format, and then imported the model into Geomagic
Studio software. This software allows for the editing Fig.9. Mesh with damaged areas identified in red.
of the point-clouds, of the generated mesh and
provides, editing functions of advanced surfacing, in
addition to its accurate functions of processing 3D
data.
The Mesh Doctor is an automatic improvement of
polygon mesh. It is generally recommended to use
the Mesh Doctor after importing a polygonal model.
Fig.10. Measurement of the main entrance by two
points in Photoscan.

4. Validation of the results obtained

with the photogrammetric survey.
3D photogrammetry is affected by inherent errors
([21],[20]) arising from processing algorithms and by
those typical of photography.
Picture quality, for instance, is limited by sensor
resolution, lens optical deformation and picture
prospective. The latter requiring straightening of the
image with consequent dimensional and parallax
errors.
Other well-known error source arise from image
mosaicing [12]. To assess the accuracy of the model,
a set of repeated measurements were taken on Fig. 11. Measurement from Photoscan.
selected part of the building to be compared with
reference measurement taken with precision
instruments (TOPCON Electronic Total Station,
GTS-312).
The need to perform multiple measurements on
the 3D model derives from the uncertainty offered by
the pointing system for the measurements inside the
photogrammetric software.
The results of the comparisons for some of the
elements of the building (see Figs.10-11), along with
the errors in term of difference and percentage
difference are given in Tables 1-8.
Finally, Table 9 shows the mean and standard
deviation on the measured element of the structure.
We can note that error between the measurements
obtained with the survey and the 3D model obtained
from Photoscan, after the scaling, is always less than
1%.
Fig.12. Planimetric survey and measured elements
(TOPCON Electronic Total Station, GTS-312).
 Table 4. Comparison between survey and Photoscan
measurements – EF element

Survey Photoscan
n  [m] 
[m] [m]
1 5,829 5,801 0,0279 0,48%
2 5,829 5,795 0,0335 0,57%
3 5,829 5,858 -0,0295 -0,51%
4 5,829 5,862 -0,0336 -0,58%
5 5,829 5,804 0,0250 0,43%
6 5,829 5,794 0,0344 0,59%

Fig.13. Perspective drawing

Table 5. Comparison between survey and Photoscan

Table 1. Comparison between survey and Photoscan measurements – BE element
measurements – AB element
Survey Photoscan
Survey Photoscan n  [m] 
n  [m]  [m] [m]
[m] [m]
1 13,768 13,713 0,0555 0,40%
1 5,463 5,435 0,0282 0,52%
2 13,768 13,830 -0,0615 -0,45%
2 5,463 5,429 0,0343 0,63% 3 13,768 13,834 -0,0653 -0,47%
3 5,463 5,496 -0,0332 -0,61% 4 13,768 13,710 0,0585 0,42%
4 5,463 5,434 0,0290 0,53% 5 13,768 13,715 0,0529 0,38%
5 5,463 5,501 -0,0383 -0,70% 6 13,768 13,700 0,0682 0,50%
6 5,463 5,431 0,0315 0,58%

Table 2. Comparison between survey and Photoscan Table 6. Comparison between survey and Photoscan
measurements – BC element measurements – IJ element

Survey Photoscan Survey Photoscan

n  [m]  n  [m] 
[m] [m] [m] [m]
1 5,531 5,564 -0,0324 -0,59% 1 16,285 16,344 -0,0588 -0,36%
2 5,531 5,496 -0,0351 -0,64% 2 16,285 16,350 -0,0651 -0,40%
3 5,531 5,555 0,0234 0,42% 3 16,285 16,237 0,0477 0,29%
4 5,531 5,502 0,0290 0,52% 4 16,285 16,342 -0,0568 -0,35%
5 5,531 5,563 -0,0313 -0,57% 5 16,285 16,213 0,0716 0,44%
6 5,531 5,494 0,0375 0,68% 6 16,285 16,216 0,0686 0,42%

Table 3. Comparison between survey and Photoscan Table 7. Comparison between survey and Photoscan
measurements – CD element measurements – H element

Survey Photoscan Survey Photoscan

n  [m]  n  [m] 
[m] [m] [m] [m]
1 5,563 5,598 -0,0345 -0,62% 1 14,843 14,777 0,0665 0,45%
2 5,563 5,534 0,0289 0,52% 2 14,843 14,902 -0,0593 -0,40%
3 5,563 5,533 0,0299 0,54% 3 14,843 14,913 -0,0782 -0,53%
4 5,563 5,592 -0,0290 -0,52% 4 14,843 14,765 0,0699 0,47%
5 5,563 5,595 -0,0321 -0,58% 5 14,843 14,768 0,0747 0,50%
6 5,563 5,526 0,0367 0,66% 6 14,843 14,922 -0,0786 -0,53%
Table 8. Comparison between survey and Photoscan Table 10. DXF entities - SAP2000 objects
measurements – Main entrance correspondence table

Survey Photoscan
n  [m]  DXF entity SAP2000 object Color
[m] [m]
1 0,892 0,886 0,0064 0,72% Point Joint Yellow
2 0,892 0,883 0,0087 0,97%
3 0,892 0,885 0,0072 0,81% Line Frame Yellow
4 0,892 0,898 -0,0056 -0,63% Point Link (one-point) Green
5 0,892 0,897 -0,0048 -0,53%
Line Link (two-point) Green
6 0,892 0,885 0,0073 0,82%
3D Face Shell Red
Polygon
Solid White
Mesh
Table 9. Mean value µ and deviation σ of the
measurements A snapshot of the finite element model after the
import procedure in SAP2000 is given in Fig.14 with
some properties (number of degrees of freedom,
Element µ [m] σ [m]
number of dynamic degrees of freedom and number
A-B 0,0086 0,0314 of triangular shells) included in Table 11.
B-C 0,0024 0,0317
D-E 0,0000 0,0320 Table 11. Properties of the imported SAP2000 model
E-F 0,0096 0,0293
B-E 0,0180 0,0578 Number of degrees of freedom 395.682
I-J 0,0012 0,0620 Number of mass degrees of freedom 197.841
H 0,0019 0,0715 Number of shell elements 145.352
Main
entrance 0,0032 0,0060 For the assessment of the finite element model
generated by the proposed procedure the structural
analysis under seismic design load is performed.
It should be noted that an accurate structural
analysis would require the estimation of the
5. Structural analysis. mechanical properties of the materials through
At this stage, the model of the castle is first imported experimental tests (see [15] and [16]). However, for
into AutoCAD to be rearranged. the purpose of this preliminary analysis phase,
Then the model is exported in Drawing Exchange mechanical properties suggested by the Italian
Format (DXF). building code for similar masonry typologies
DXF entities should be organized by assigning to (NTC2008, [7]) are assigned to shell elements (see
a different layer each DXF entity to matching objects Table 12). Thickness of masonry walls varies from
of the finite element analysis software. 2.5 m at the base to a minimum of 1.5m at the roof.
The finite element software Sap2000 (SAP2000 The software automatically calculates the total mass
CSI inc. Berkeley, Cal, US [11]) was selected for the of the structure, equal to 2396 t.
analysis and the correspondence-table between DXF
entities and Sap2000 objects are shown in Table 10. Table 12. Mechanical properties of masonry
DXF files are finally imported into SAP2000 elements
using the software specific import procedure.
Type of 𝒇𝒎 𝝉𝟎 𝑬 𝑮 𝜸
stone
(𝑵⁄ 𝟐 ) (𝑵⁄ 𝟐 ) (𝑵⁄ ) (𝑵⁄ ) (𝑲𝑵⁄ 𝟑 )
masonry 𝒄𝒎 𝒄𝒎 𝒎𝒎𝟐 𝒎𝒎𝟐 𝒎
Squared
Coursed 800 12.0 3200 940 22
blocks
𝑓𝑚 = compressive strength, 𝜏0=shear strength, 𝐸=elastic
modulus, 𝐺 = Shear modulus,
𝛄 = unit weight
 Table 13. Design parameters for the evaluation of the
acceleration response spectrum.

Nominal life VN [years] 50

Functional type CU [-] III
Design reference life VR [years] 75

Results of the modal analysis (see Table 14) show

that the first (Fig.16) and the second mode (Fig.17)
are mainly translational along the weak-axis, i.e.
orthogonal to the line connecting the two circular
towers, and the strong-axis of the structure,
respectively. The third mode (Fig.18) exhibits,
instead, a relevant rotational component.
Principal stresses from gravity load analysis are
shown in Fig.19. These results show limited and
reasonable stress concentration, giving a good
feedback on the quality of the finite element model.
Furthermore, compressive stresses at the base of
the structure, with values varying between 3.15
daNcm-2 and 4.54 daNcm-2, are quite close to the
Fig.14. Imported model in Sap2000. mean vertical stress component at the base, simply
obtainable by taking the ratio of the total weight over
Seismic analysis to special provision for existing the structure base area, equal to 3.74 kgcm-2.
structures of the Italian building code is performed.
Site-specific response-spectrum for the considered
life-safety limit state is evaluated using parameters Table 14. Mass participation ratio and their
included in Table 13 with the resulting elastic summation along the degrees of freedom X, Y and
response spectrum shown in Fig. 15. rotational Z for the first 5 modes.

Sum Sum Sum

Mass Mass Mass
Mode Period Mass Mass Mass
ratio ratio ratio
n. (s) ratio ratio ratio
X Y RZ
X Y RZ
1 0.211 0.203 0.390 0.203 0.390 0.004 0.004

2 0.156 0.238 0.113 0.441 0.502 0.003 0.007

3 0.143 0.029 0.058 0.471 0.560 0.147 0.154

4 0.141 0.157 0.060 0.628 0.620 0.107 0.261

5 0.092 0.000 0.000 0.628 0.620 0.418 0.679

Moreover, an equivalent static analysis is

performed combining the seismic input in one
direction, i.e. X or Y, with the 30% of the orthogonal
one. Maximum and minimum principal stress
components for the shell elements subjecting the
model to the mentioned lateral forces are shown in
Fig.20. Shear stress values comprised between 1.92
daNcm-2 and 2.69 daNcm-2, see Fig.21, show that
there are not anomalous stress concentrations due to
poor meshing. Other stress components are not
included here for brevity.
Fig.15. Elastic response spectrum at the life-safety
limit state.
 Finally, a preliminary stress-check under seismic
loads is performed here by comparing design stress
components at the base of the structure (seismic
demand) against shear stress capacity. The design
shear capacity Vt is estimated by the following
equation (see [7])

𝑉𝑡 1.5∙𝜏0𝑑 𝜎 0
= √1 + 1.5∙𝜏 , (1)
𝑙∙𝑡 𝑏 0𝑑

where
l is the length of the wall;
t is the thickness of the wall;
𝜎0 is the vertical mean stress component divided
to the confidence factor;
b is a coefficient (1 ≤ 𝑏 ≤ 1.5) equal to h/l,
where h is the height of the wall;
𝜏0𝑑 design shear strength (see Table 12).
(a)
Table 15. Demand to Capacity ratio under seismic
load at the base of the structure.

𝑽𝒕
𝝉𝟎 𝒍∙𝒕 𝑫/𝑪
(daN/cm2) (daN/cm2)

1,92 ÷ 2,69 2,95 0,65 ÷ 0,91

𝑉
𝜏0 = seismic shear stress component; 𝑡 =design
𝑙∙𝑡
shear capacity; D/C=demand to capacity ratio

The values of the shear stress 𝜏0 and vertical stress

component 𝜎0 are given in Table 15. The results of
this analysis show a Demand to Capacity ratio
varying between 0,65 and 0,91 at the base of the
structure which is lower than the limiting value of
2,25 suggested by the Italian code provisions.
Moreover, it appears that at some locations stress
concentrations, due to rapid change in mesh
geometry, are still noticeable. (b)

Fig.16. First mode T= 0.211 sec (Grey = undeformed

shape): (a) top view and (b) lateral view.
 (a)
(a)

(b) (b)

Fig.17. Second mode T= 0.156 sec (Grey = Fig.18. Third mode T= 0.143 sec (Grey =
undeformed shape): (a) top view and (b) lateral view. undeformed shape): (a) top view and (b) lateral view.
 (a) (a)

(b) (b)

Fig.19. Minimum (a) and maximum (b) principal Fig.20. Minimum (a) and maximum (b) principal
stress (kg/cm2) for the gravity load combination. stress (kg/cm2) for the seismic load combination in
the x direction plus 30% in y direction.
 However, even though promising, some issues
arise from the great number of 3D-faces produced by
the automated generation algorithms, requiring
further post-processing efforts for the reduction of
the finite elements mesh to a more manageable size,
and from to the inability to differentiate, during the
acquisition phase, the mechanical properties of
different part of the model. Furthermore, the
generated mesh still needs further refinement to fix
various mesh errors.

References:
[1] A. Bandiera, J.A. Beraldin, M. Gaiani, Nascita
ed utilizzo delle tecniche digitali di 3D imaging,
modellazione e visualizzazione per l’architettura
e i beni culturali, Ikhnos, 2011, pp. 81-134.
[2] L. Guidi, Acquisizione 3D e modellazione
poligonale. Edizione Mondadori, Milano, 2010.
(a)
[3] L. Barazzetti, G. Forlani, F. Remondino, R.
Roncella, M. Scaioni, Experience and
achievements in automated image sequence
orientation for close-range photogrammetric
projects, Proc. of SPIE Optics+Photonics, Vol.
8085, 23-26 May, Munich, Germany, 2011.
[4] V. Barrile, G.M. Meduri, G. Bilotta,
Comparison between Two Methods for
Monitoring Deformation with Laser
Scanner, WSEAS Transactions on Signal
Processing, Vol.10, 2014, pp.497-503.
[5] V. Barrile, G.M. Meduri, G. Bilotta, Laser
scanner technology for complex surveying
structures, WSEAS Transactions on Signal
Processing, No.3, Vol.7, 2011, pp. 65-74.
[6] V. Barrile, G.M. Meduri, G. Bilotta, Laser
scanner surveying techniques aiming to the
study and the spreading of recent architectural
structures, Recent Advances in Signal and
(b) Systems, Proceedings of the 9th WSEAS
International Conference on Signal, Speech and
Fig.21. Shear stress (kg/cm2) for the seismic load Image Processing (SSIP '09), 2009, pp.92-95.
combination in the x direction plus 30% in y [7] NTC 2008. D.M. del Ministero delle
direction. Infrastrutture e dei Trasporti del 14 gennaio
2008. Nuove Norme Tecniche per le
Costruzioni.
4 Concluding remarks [8] Agisoft PhotoScan v.1.1
The techniques used in this work have shown the (http://www.agisoft.com/).
feasibility of low-cost photogrammetry for survey [9] Geomagic Studio 2013
and structural modeling purposes of a relevant (http://www.geomagic.com/).
construction. The adopted procedure, even though [10] Autodesk Autocad 2016
not completely automated, is capable to produce - (http://www.autodesk.it/).
contactless and without the need of many known [11] CSI Sap2000 v.17.1 (http://www.csi-
reference points - a satisfactory 3D model from italia.eu/).
digital images with a degree of accuracy totally [12] R. B. Inampudi, Image mosaicing, IGARSS
beyond the reach of ordinary survey techniques. '98. Sensing and Managing the Environment.
 1998 IEEE International Geoscience and Computers & Structures, Elsevier Ltd, 2011,
Remote Sensing Symposium Proceedings. (Cat. No.89, pp. 1615-1627.
No.98CH36174), Vol.5, 1998, pp. 2363-2365. [18] A. Bhatla, S.Y. Choe, O. Fierro, F. Leite,
[13] E. D’Amore, S. Trovato. Experimental Evaluation of accuracy of as-built 3D modeling
analysis for identification of bridges structural from photos taken by handheld digital cameras,
damage Operational using Modal Analysis Automation in Construction, Elsevier B.V.,
based methods, Proceedings of 15th World 2012, No.28, pp. 116-127
Conference on Earthquake Engineering (15 [19] P. Arias, J. Herráez, H. Lorenzo, C. Ordónez,
WCEE), 2012. Control of structural problems in cultural
[14] A. Carpinteri, G. Lacidogna, Damage heritage monuments using close-range
Monitoring of an Historical Masonry Building photogrammetry and computer methods,
by the Acoustic Emission Technique, Materials Computers & Structures, 2005, No.83, pp.
and Structures, 2007, No.39, pp. 161-167. 1754-1766.
[15] L. Rovero, V. Alecci, J. Mechelli, U. [20] T. Läbe, W. Förstner, Geometric stability of
Tonietti, M. De Stefano, Masonry walls with low-cost digital consumer cameras, Proceedings
irregular texture of L’Aquila (Italy) seismic of the 20th ISPRS Congress, Istanbul, Turkey,
area: validation of a method for the evaluation 2004, pp. 528-535.
of masonry quality, Materials and Structures, [21] X. Li, Photogrammetric investigation into
Springer Netherlands, 2015, pp. 1-18. low-resolution digital camera systems, Thesis
[16] M. Tomaževič, Shear resistance of masonry (Ph. D.), Dept. of Geodesy and Geomatics
walls and Eurocode 6: shear versus tensile Engineering, University of New Brunswick,
strength of masonry, Materials and Structures, 1999.
No.42, 2009, pp. 889-907. [22] V. Barrile, G.M. Meduri, G. Bilotta,
[17] I. Lubowiecka, P. Arias, B. Riveiro, M. Experimentations and Integrated Applications
Solla, Multidisciplinary approach to the Laser Scanner/GPS for Automated Surveys,
assessment of historic structures based on the WSEAS Transactions on Signal Processing,
case of a masonry bridge in Galicia (Spain), 2014, Vol.10, pp. 471-480.