ACTA IMEKO
ISSN: 2221-870X
March 2021, Volume 10, Number 1, 84 - 97
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 84
Digital surveying and 3D modelling structural shape pipelines
for instability monitoring in historical buildings: a strategy of
versatile mesh models for ruined and endangered heritage
Sandro Parrinello1, Raffaella De Marco1
1 DICAR – Department of Civil Engineering and Architecture, University of Pavia, via Ferrata 3, 27100 Pavia, Italy
Section: RESEARCH PAPER
Keywords: Endangered Heritage; reality-based modelling; digital survey; structural modelling; Cultural Heritage Routes
Citation: Sandro Parrinello, Raffaella De Marco, Digital surveying and 3D modelling structural shape pipelines for instability monitoring in historical buildings:
a strategy of versatile mesh models for ruined and endangered heritage, Acta IMEKO, vol. 10, no. 1, article 12, March 2021, identifier: IMEKO-ACTA-
10 (2021)-01-12
Editor: Eulalia Balestrieri, University of Sannio, Italy
Received June 1, 2020; In final form November 23, 2020; Published March 2021
Copyright: This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie
grant agreement No 821870.
Corresponding author: Raffaella De Marco, e-mail: raffaella.demarco@unipv.it
1. INTRODUCTION: THE DOCUMENTATION OF
ENDANGERED HERITAGE
Both the analysis of cultural heritage and the development of
communitarian guidelines for its protection [1] are determining a
growing scientific approach to European and worldwide heritage
sites [2]. The variety of built heritage (historical buildings,
monuments, historical centres, sites, and territorial landscapes)
requires a wider field of knowledge and intervention in terms of
both physical structures and cultural policy. This leads to the
attendant difficulties in protection and preservation due to a
fragmented reality of separated protocols and documentation,
which results in difficulties in the sharing of information and data
integration, thus hindering the entire approach to an intervention
programme. This is especially the case for the so-called
‘endangered heritage’ sites (World Heritage Convention, List of
World Heritage in Danger 1972), the class of built heritage
particularly affected by proven or potential threats that pose a
high level of risk for its preservation [3].
In fact, the revision of the sites officially listed as endangered
heritage has been achieved in terms of the conditions of
relevance and correspondence with the specified risk criteria and
indicators (territorial, political, or social dangers). However, the
statistics in terms of monitoring parameters and protocols
remain non-uniform, highlighting a lack of scientific and
technical documentation and a lack of connection between
intervention activities and ‘reality-based’ updated analysis. The
framework highlights the coexistence of a dual type of
emergency issue related to built heritage, with the classification
ABSTRACT
Cultural heritage and the attendant variety of built heritage demands a scientific approach from European committees: one related to
the difficulties in its protection and management. This is primarily due to the lack of emergency protocols related to the structural
knowledge and documentation pertaining to architecture and its ruins, specifically in terms of the goals of protection and intervention
for endangered heritage affected by mechanical instabilities. Here, we focus on a rapid and reliable structural documentation pipeline
for application to historical built heritage, and we introduce a case study of the Church of the Annunciation in Pokcha, Russia, while we
also review the incorporation of integrated 3D survey products into reality-based models. This practice increases the possibility of
systematising data through methodological phases and controlling the quality of numerical components into 3D polygonal meshes, with
millimetric levels of detail and triangulation through the integration of terrestrial laser scanner and unmanned aerial vehicle survey data.
These models are aimed at emphasising morphological qualities related to structural behaviour, thus highlighting areas of deformation
and instability of the architectural system for analysis via computational platforms in view of obtaining information related to tensional
behaviour and emergency risks.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 85
of the site as ‘endangered’ for the stability and integrity of its
physical
structure on the one hand, and the growing demand for specific
parameters and protocols for the analysis and quantification of
the emergency value of the building, thus determining the
effective intervention.
Thus, there is a growing demand for the identification of such
sites [4], both in geographical and typological terms, expanding
the dissemination and application of appropriate monitoring and
knowledge practices prior to intervention [4], with the aim of
triggering a process of safeguarding policies [6].
The topic of implementing safeguarding methodologies for
existing heritage is receiving a great deal of attention among
researchers, with new forms of digital products emerging in
terms of a survey-based phase of digital documentation related
to cultural heritage, which is crucial for an accurate and complete
understanding of the characteristics and parameters of the units
and contexts of built heritage. This is being combined with
computer-based cognitive and interactive models, with the
elaboration of 3D digital products for investigations and
simulations related to shapes and structures.
These new representation systems are resulting in new
expectations related to digital communication, changing the
objectives and constantly renewing the demand in analytical
terms for cognitive requirements, in response to the
requirements linked to the computational nature of interaction
within the models themselves, which are now capable of
providing both quantitative and qualitative answers. The
difficulties in the development of the reliable diagnosis of
historical structures can be realigned with the need for new
methods of analysis that can exploit the computational
opportunities through specific methodologies and cognitive
practices. These practices relate to the visual and graphic aspects
of the documentation pertaining to architecture [7], reliably
representing the present state of stability and linking it to the
constructive and safeguarding rules.
2. HISTORICAL STRUCTURES AND INSTABILITIES IN THE
DIGITAL DOCUMENTATION PROCEDURES
Historical buildings are traditionally founded on and remain
linked to their original structural system, from the construction
phase to later procedures. The demolition of bearing systems and
elevations, the reconfiguration of horizontal levels and vaults,
and the construction of expansion blocks thus enrich the
mechanical experience of the monument (Figure 1).
The consideration of structural diagnosis applied to historical
built heritage in terms of the knowledge of the stress behaviours
and the prevision of damage mechanisms plays a central role in
the development of the documentation protocols for
safeguarding. A review of the risks, the on-going kinematics, and
the priorities of stability related to endangered buildings [8]
places the focus on the aspect of robustness, that is, the capacity
of the building and its elements to withstand a certain level of
stress imposed by a combination of degradation and the changes
in both the materials and the environments [9]. Thus, considering
the deep impact that stress-altering phenomena have on the
geometrical and material aspects of buildings (e.g. deformations,
drifts, cracks) is part of a comprehensive diagnosis that can be
achieved through an investigation into the structural shape [10].
As such, it is possible to focus the evaluation of the stress
instabilities on the target of a ‘static morphology' that reflects the
mechanical and stress traces left by the adaptation schemes of
the system during its physical history. It also highlights the
stability principles related to the class and type of resistant
geometries (Figure 2).
Nowadays, the analysis of the structural aspects of historical
buildings involves simplified typological schemes transformed
into numerical structures, selected to be mechanically
Figure 1. Examples of sites along the Upper Kama route characterised by on-
going risks of abandonment and decay, proving the basis for a classification
as endangered heritage. From the top, Church of the Exhaltation of the Holy
Cross in Bondjug, Rubezhskaya Church in Usolye, Church of St. Nicholas in
Uzhginskay, Church of Paraskeva Friday in Saltanovo.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 86
controllable in defined levels of physical regularisation and
abstraction [11]. However, the confidence in the standardisation
of structural computing has led to underestimating the direct
‘experience’ of the building, visible in the singular imperfection
of its shape.
In fact, structural documentation must be both 'comparative',
with reference to the archetypes, and 'specific', highlighting the
accurate variation in formal detail. The testing of specific
intervention procedures for existing components aimed at
assessing their instability performance is highlighting the need
for more precise models for the creation of reliable simulations
[12]. At the same time, the area of digital documentation has
developed to include morphometric 3D databases with more
dense information, which allows for measuring and analysing
space in terms that is closer to the continuous form [13]. The
relationships between the shape and mechanics of structural
systems can determine new descriptive products through reality-
based numeric surfaces in terms of high-poly mesh models,
which can be morphologically assessed to extract specific
quantitative parameters that can guide the mechanical
interpretation. Thus, the reality-based mesh model, the numerical
component of which reliably determines and characterises the
geometric aspects, can become an adaptable tool for the
management of the computational and analytical processes
related to robustness.
These considerations are encouraging strategies of data
integration for the comprehensive adoption of structural reality-
based models related to architectural heritage. On a scientific
level, the experimentation of reality-based models for structural
diagnosis will result in a multidisciplinary and implementable
methodology, one capable of guaranteeing a standardised
product, that is, the polygonal mesh model, involving different
levels of detail and integration for the management of the
existing built heritage. This type of model is aimed at determining
both emergency and long term interventions, in the calibration
of their procedural computing in both a scientific and practical-
operational direction through the capacity for the shapes to be
transformed into morphological and computational platforms of
analysis, such as finite element analysis (FEA) platforms. Here,
the methodological process must be as fast, as extendable, and as
replicable as possible to facilitate interchange and to allow for a
complete knowledge and management capacity through 3D
models related to buildings that are in a state of emergency.
3. AN EXPERIMENTAL CASE STUDY: THE CHURCH OF THE
ANNUNCIATION IN POKCHA (RUSSIA)
3.1. The site and its historical/structural characterisation
The constructive features of built structures relate to the
technological and material traditions pertaining to specific
continents, where the stylistic influences and source availability
of certain elements have determined the specific characteristics
of the architectural shape. This is especially the case with the built
sites and monuments along the European Cultural Heritage
Routes, the architectural value of which is strongly permeated by
specific structural typologies. These architectural structures have
thus acquired their own identity, characterisation, and specificity
over time, evolving specific spatial distributions and constructive
apparatuses in their historical development and functional
change. Resilient structures combine a wide framework of
morphological and static relationships that are reflected in their
own state of conservation and which influence the robustness-
related approach to their restoration and preservation for
security and stability.
The case of Upper Kama, Russia, already the subject of
international research related to the digital documentation of the
complexes [14], presents an emblematic example of the Cultural
Heritage Route and the coexistence of historical structures and
various phenomena of instability caused by the functional
experience of the building, reflecting a complex framework of
diagnosis for the safety and integrity of the various built sites. In
these terms, the Upper Kama sites form a rich morphological
abacus of technological modules and elements of structure, along
with the related pathologies of degradation and conservation [15]
[16].
The cultural tradition of the Upper Kama settlements, more
stabilised by commercial requirements than most villages of the
Russian countryside, has determined a global phenomenon of
transformation regarding the built structural systems, from the
typical wooden architecture to the constructive solutions of brick
and stone masonry. Therefore, these architectural structures are
characterised by a morphological configuration involving
multiple expansion blocks. During the Soviet period, the demand
for the conversion of infrastructures for energy or food
production resulted in the decline of these religious buildings.
The consequent interventions and the alteration of the
architectural structures, forcibly adapted to the new functions,
led to a prevalent abandonment characterised by eventual ruin
and collapse.
Figure 2. The state of neglect and the lack of conservation present critical situations of widespread instability among the masonry systems, clearly exposing
the structure to the damage affecting its elements. Robustness instabilities in the structural complex of the Church of the Transfiguration, Pyskor, defining the
increasing frameworks of damage that will destroy these buildings without specific lines of intervention.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 87
The case of Blagoveshchenskaya Church, or the Church of
the Annunciation of the Blessed Virgin, in the village of Pokcha,
represents a key monumental site of the Cherdyn district, which
synthesises various different historical-functional evolutions in
the stratification of its structures and walls, which today lie in
ruin [17], Figure 3.
The original wooden complex was replaced in 1785 with a
new structure composed of stone and brick masonry, subdivided
into multiple structural blocks, which include the main body,
with a quadrilateral planimetry, constituting a nucleus for the
refectory, the southern and northern chapels, the bell tower, and
the entrance narthex. In 1910, a reconstruction intervention
resulted in various structural modifications, especially in terms of
the bell tower, which was entirely replaced, and the eastern
section of the central vault and the altar, reconstructed with the
insertion of a five-headed chapter. Meanwhile, the interiors
formed in plastered stone, with paintings and ornaments dating
from 1870, remain largely preserved. The face of the building,
which features an additional red brick facing, contributes to the
strengthening of the external envelope and allows for the
possibility of inserting additional devices of tension resistance
into the stratified walls.
The history of the site involves attempted restoration works
dating from 1920 to the complete abandonment in 1940 and to
the re-conversion into a power central. The energy issues related
to the new function resulted in the partial collapse of the main
pavilion vault and the bell tower roof in the 1990s, following
repeated lightning strikes attracted by the electrical system.
Following this extensive damage, the church was excluded from
the list of architectural monuments of interest, precluding any
new interventions or restoration works, and leaving the site to
suffer virtual collapse.
3.2. The current state of the structural ruins
In 2018, the architectural complex was in a clear state of
neglected conservation. The wooden roofs had been almost
entirely destroyed, and the main masonry structure was damaged
in various areas, particularly in the main vaulted system.
Meanwhile, the vaulted structure of the spans, the loading on
the pillars based on a counter-balance scheme involving vault–
vault and vault–support relationships, has been distorted by the
collapse of the secondary vaults. In terms of the large central
pavilion vault, only around half of its total span has been
preserved, with the head brick structure left visible along the
detachment edge. Both the brick vaults and the wooden roof
have collapsed into the central environment of the nave, with the
debris submerging part of the loading pilasters of the vaults,
which cannot be inspected in their ground basement and have
been covered by soil and vegetation, creating a natural hill that
both reduces the access and is the cause of the degradation of
the preserved supports. The connection with the bell tower, once
provided by the central nave through the gallery and the
refectory, has been demolished and thus prevents the direct
documentation of the state of conservation of the elevated
structures.
Following the collapse of the roof, the complex was deprived
of the main source of protection from atmospheric agents,
particularly pertinent during the winter months, and has
therefore suffered a rapid degradation of the remaining portions,
affected year on year by localised collapses. The site is also totally
lacking in control services to prevent access to both people and
animals, who often occupy it and damage the spaces and
structures. Here, the narthex environments have been
particularly affected by the frequent presence of herds in
transhumance, often housed by the shepherds inside the church
during the summer season (Figure 4).
3.3. Objectives and targets of the documentation process
The objective of the documentation of the
Blagoveshchenskaya Church, in addition to the wider mapping
and reconnaissance approach for the sites in the Upper Kama
area, is to encourage a rapid but reliable diagnosis of its structural
crisis. This is aimed at directing the organisation of possible
intervention operations, and to safely preserve and promote the
historical ruins. While the local administration is gradually
recognising the historical and cultural value of these sites along
the Kama, the ruined conditions present a clear disadvantage for
the commitment, both technical and economic, to implementing
conservation procedures. In this sense, rapid survey and
quantification methods for recognising and evaluating the
preserved structural systems and the corresponding risk
constitute an excellent preliminary tool for accurately
determining the qualitative and quantitative planning of
interventions that could halt the exponential cycle of damage.
The documentation of the current state of conservation of the
structure of the Blagoveshchenskaya Church has thus
highlighted the need for structural diagnosis using 3D digital
survey methods. This involves applying specific integrated
approaches for the acquisition and modelling of the structural
shape, one that ensures a descriptive quality of its resistant
architectural structure through the data collected through a
Figure 3. The complex of Blagoveshchenskaya Church in Pokcha. The original
architectural structure of the church before the revolution in 1956 (above)
and the ruins as of 2018 (below).
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 88
practical and fast on-site survey. The required knowledge is
largely related to the shape alteration of the structural ruins,
ensuring that it can be comprehensively and reliably documented
and allowing for analysing the degree of safety and determining
the recovery intervention in relation to the overall mechanical
scheme.
4. STRATEGIES OF ANALYSIS FOR THE DIGITISATION OF THE
RUINED SHAPE
The documentation of the state of conservation of
Blagoveshchenskaya Church as of 2018 involved experimenting
with integrated approaches in digital surveying to focus on the
static instabilities of the ruins for recovery purposes [18]. Here,
the morphological analysis of masonry structures was organised
during the acquisition processes, semantising the spatial
constructive ruins and linking them to the global volumetric
macrosystem. Furthermore, the internal inspection of the
masonry sections in terms of the fractured or collapsed portions
has allowed for the integration of information related to the
structural envelope, reliably supporting the digitisation
procedure [19].
The ruined shape, the central element of the analysis process,
has thus become the basis of direct knowledge, supported by the
repertoire of historical/constructive analysis data, with the dual
purpose of completing the technical stratigraphic knowledge
framework and revising it using integrated research in relation to
the behavioural diagnosis of the structural components. The
ruined complex of Blagoveshchenskaya Church, for reasons
attributable to the collapse of the structures and the lack of site
regulation, has been deprived of the original spatial design of the
wall system. The absence of the main masonry portions hinders
the analysis of the current structure in terms of creating defined
schemes of interpretation in relation to the architectural
instabilities, and this indicates that the specific evaluation of the
Figure 4. Photographic gallery of the Blagoveshchenskaya Church in Pokcha.
On the left, the state of conservation in 2018, the overall structural system
and the specific static units that characterise the statics of the entire
complex. On the right, evolution of the damage to the main vault from 2006
to 2019, which increased rapidly due to the lack of intervention. Details of
the constructive system that influences the robustness evaluation of the
complex. Indication of interventions for the functional conversion into a
power central, stratigraphic technology of the masonry walls and distribution
of a steal structure inside the masonry.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 89
reality-based digital shape is a crucial step for a reliable diagnosis
of the static robustness [20].
The ruined complex has been subjected to an integrated,
extensive, range- and image-based documentation campaign.
The attendant approach involved calibrating methodologies of
acquisition and representation in terms of the morphological
properties of the structural elements, especially in terms of the
correspondence and integration of spatial information. The
dimensions of the structure (specifically the elevation of the
damaged portions of the main vault at over 15 m in height) have
oriented the morphometric digitisation strategy toward the use
of multi-instrumental close-range measuring systems, aimed at
guaranteeing a comprehensive coverage and a high formal
detailed resolution comparable to the complexly shaped
geometry of the masonry surfaces and the typological damage of
the masonries (widely covering the order of dimensions from 5
to 10 cm).
At the same time, the colorimetric information has proven to
be a key aspect in the diagnostics of the structures, evaluating the
pathologies of mechanical influence on the masonry and
plastered portions, described through tonal variations in the
surface colours.
The on-site campaign was organised on a double acquisition
level, at the ground level using a terrestrial laser scanner (TLS)
and at an aerial level using drones for photogrammetry, with the
aim of obtaining comprehensive morphological data on the
elevations of the structure (over 15 m) in a single 360 ° point
cloud database [21].
4.1. Terrestrial laser scanning acquisition campaign and products
In the acquisition using a TLS FARO Focus S150, the
integrated photographic camera was also enabled, ensuring a red,
green, blue (RGB) information quality for each point (x, y, z) that
is suitable for surface mapping. Each scan was performed in
medium or high mode according to the distance of the shooting
range from the object and provided an additional acquisition
time of 4.00 minutes to allow for the shooting of 16 frames for
each camera location (coverage of 360 ° × 320 ° for the scan
angle). Prior to each photographic shot, the instrument was set
with an average balance of colour and lighting while considering
only the horizontal level for the balance of light conditions in the
outdoor scans, or the entire panoramic space for the interior
scans. These measures significantly affected the acquisition
campaign in terms of duplicating the scanning times but were still
more advantageous for the documentation purposes than a
parallel photographic campaign from the ground, which requires
further actions to integrate the photographic material into the
spatial database. A total of 73 TLS scans were realised to collect
all external surfaces and to spatially connect the complex
distribution of the internal environments. The scans were
performed at an approximate 2-mm laser spot spacing up to a
height of 5 m, and close to 5 mm in the upper surfaces. The
quality of TLS acquisition for the bell tower and the central dome
was ensured by the presence of the inner natural hill over the
ruins of the roof, which allowed for a higher instrument
positioning above ground level. The TLS metric survey
conducted from the ground guaranteed a coverage of 80 % of
the morphological data related to the structure (Figure 5).
The processing of the seven source scans on the vaulted
structure provided an alignment procedure to obtain the unified
TLS point cloud. Rotation and translation matrixes were then
applied in semi-automatic mode via the cloud software
(SCENE), using a cloud-to-cloud alignment set with values of
0.05 m in subsampling and n° 150 iterations for the alignment
algorithm. The resulting values of alignment characterising the
final integrated database are presented in Table 1.
4.2. Unmanned aerial vehicle acquisition campaign
At the same time, the aerial photogrammetric acquisition was
developed using an unmanned aerial vehicle (UAV, DJI
Phantom 4 Pro), with the flight plan focused on the points of
interest of the architectural complex. In fact, the flight plan
mission was set from the top centre of the complex at a level of
50 m from the ground in the mode ‘point of interest’. Here, a
photogrammetric campaign was conducted using the aerial
camera with a conical acquisition surrounding the monumental
complex, descending to a height of 10 m above ground and
developing 329 shots in 10 min of flight. The guaranteed camera
angle and the adjustable inclination of the gimbal allowed for
total coverage of the top surfaces, with a quality of photographic
data suitable for the clear photogrammetric reconstruction of the
edges of the components (21 MPixel). In addition, specific flight
plans for the tower blocks and the central vault were planned
with greater photographic detail, largely due to the close distance
of shooting. The GPS coordinates defined by the UAV for each
shot better addressed the alignment algorithms, set to ‘high’
precision, simplifying the alignment of the photogrammetric
structure from motion (SfM) reconstruction. In terms of the
overall SfM point cloud, there was an error of 0.004 pixels due
to the camera alignment (Figure 6).
4.3. Post-processing data and products
The GPS information in the source data obtained via the
acquisition campaign provided a preliminary overlapping
between the TLS- and UAV-acquired point clouds, both in terms
of vertical and horizontal built surfaces, which was used to
optimise the referencing in an integrated sparse database of
morphometric characters.
To optimise the GPS data, the alignment procedure for the
TLS and UAV data was set with reference to the TLS point
cloud, defining six markers from key morphological features that
were recorded on the UAV point cloud. The resulting values of
alignment characterising the final integrated database are
presented in Table 2. The higher errors of reference
corresponded to the markers with less camera projection in
terms of photogrammetric alignment.
The final database featured 51.889.557 and 49.853.646 points
from the TLS survey on the internal surfaces of the masonry
structure, and 2.035.911 points from the UAV survey on the
external surface and the ruins of the vaulted structure. The
average density in the overall integrated point cloud was 1
point/3 mm.
Table 1. Alignment values in cloud-to-cloud reference for TLS data.
Scan
Morpho
feature
iterations
Point error
iterations
in mm
Average point
error
in mm
Min
overlapping
in %
Scan 39
4
2.5
2.3
35.5
Scan 40
12
2.6
1.8
25.8
Scan 41
7
1.9
1.6
29.5
Scan 42
6
1.5
1.2
29.5
Scan 43
5
1.2
1.0
49.4
Scan 46
6
1.3
1.0
33.8
Scan 48
4
2.2
2.0
26.3
Cluster report
44
1.8
1.5
32.8
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 90
From the joint processing of the collected archives, useful
reference drawings, sections, and plans were produced to obtain
comprehensive knowledge of the architectural complex and to
serve as guidance for the intensive mapping of the pathologies
of degradation and instability documented in the photographic
repertoire. However, the 2D representation barely described the
quality of distribution and the volumetric development of the
main structural elements, rendered even more complex by the
ruined condition of the site and, as such, by the loss of many of
the main shapes conventionally identifiable for the classification
of the elements (intrados of the vaults, opening profiles and local
collapses, integrity of the elevation structures).
In addition, the attempt at processing for elevation maps, as
a standard analysis of deformation mechanisms using the point
Figure 5. Morphometric database from on-site TLS acquisition campaign: the single scans, completed via GPS, compass, and inclinometer information, were
referenced and registered to control the overall alignment to verify the final point cloud.
Table 2. Alignment values in TLS and UAV data morphometric reference.
Marker
Morpho feature
correspondence
Projections
Error in m
Point 1
Vault springer 1
89
0.0383
Point 2
Vault springer 2
130
0.0472
Point 3
Vault springer 3
114
0.0483
Point 4
Vault springer 4
85
0.0439
Point 5
Vault key 1
(damaged border)
87
0.0314
Point 6
Vault key 2
(damaged border)
77
0.0169
Control points
6
582
0.0392
Figure 6. Morphometric database from the on-site photogrammetric UAV acquisition campaign: the photos were geo-referenced and aligned to define the
main point cloud on the architectural complex and the surrounding landscape.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 91
cloud displacement from a reference plane [22], did not satisfy
the knowledge requirements in terms of the deformation of the
structural surfaces. The high surface processing of the structural
envelope, due internally to the collapse conditions and externally
to the decorative brick structure, resulted in difficult-to-read
colour maps, where the excessive fragmentation of the level
indicators failed to provide an overall interpretative picture of the
deformations.
5. OPPORTUNITIES FOR 3D INTEGRATED MODELLING FOR
THE DIAGNOSIS OF THE STRUCTURAL SHAPE
The aim of the structural representation was focused on the
development of a continuous mesh modelling protocol, enriched
with reality-based qualities, achieved as part of the objective of
digitally surveying detailed and imperfect shapes as confirmed by
the instability and mechanical features. These shapes, controlled
and certified within the envelope of the structural object,
necessarily introduce a need to manage the polygonal geometric
mesh that defines them, the processes of which are, however,
improved by the representation of a morphological detail that is
itself a source of diagnostic reflections on the mechanisms of
analytical systems [22] [24], Figure 7.
In this sense, the obtained 3D models not only pursue the
necessary analysis of the deviation between static surfaces but
also consolidate a certified practice that defines them both as
outputs, directly applicable for the quality of structural
instabilities, and as input data for further implementation,
specifically on computational platforms.
As such, a digital data processing was established that will
exponentially led to the exploitation of the potential of the 3D
discrete database, namely, the point cloud, and the continuous
3D object defined by vertex and edges for quantifiable
evaluations, namely, the model. The application opportunities,
from FEM platforms to information systems that enhance the
virtual fruition, can thus provide the grounding for involving the
models using common languages, thus adapting them
appropriately [25] [26], Figure 8 and Figure 9.
The developed research on reality-based mesh models has
allowed for accurately delineating the causes and effects
produced by the structural survey in the digitisation and
modelling processes related to morphometric data, validating the
compatibility and correspondence of the produced models via a
multi-instrumental comparison of the measurement practices
normally applied in the structural field. The integration of the
products of digital survey protocols applied on the site [26], from
both terrestrial and aerial image-acquisition tools, was
complementarily completed through the differentiated visual
stations, ensuring the capacity to obtain information related to
both the exterior and the interior parameters, as well as to
monitor the data related to the roof components and elevation
units. Focusing on the central pavilion vault, which was largely
destroyed during the electrical accident, the documentation,
finalised to the reinstatement of a complete structural shape,
involved the detection of the vault both from the intrados level,
also visible in terms of its constructive thickness, and the
extrados level, dominated by the ruins of the octagonal masonry
tholobate at the base of the wooden roof, Figure 10.
5.1. Modelling pipeline
The pipeline for 3D reality-based modelling was assessed in
relation to the research objectives regarding the diagnosis of the
stability of the Blagoveshchenskaya Church in Pokcha. The
Figure 7. Drawings and elaborations for the 2D description of the main
structural system of the church. The presence of noise data for elements such
as the vegetation and collapsed ruins is highlighted.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 92
choice was oriented both toward a spatial configuration of the
complex's elements, articulated on several levels and
environments, and for a correspondence of the actual state of
ruin of the building, which highlighted various critical issues to
be transposed automatically in terms of a digitised model through
non-uniform rational basis spline (NURBS) or parametric
modelling. First, the data collected from the TLS and UAV
acquisition campaigns were tested using a compatible procedure
for the integration of the processed point clouds, also supported
by common format extensions for their inclusion in the
modelling platforms. During this phase, various issues emerged
in the specific management of the vertex and polygon objects.
Corrections and uniformities of the normal vertex data
The first aspect related to the difficulty of transferring the
‘vertex normal’ information for each point of the
photogrammetric cloud acquired via the UAV on the mesh
modelling platforms. The compatible formats (mainly .ptx and
.pts) for the reverse modelling provided a different ASCII code
Figure 8. Quality of processed morpho-metric TLS and UAV data on the external surfaces of the masonry structure. The point cloud from the UAV acquisition
shows a resolution and density of data comparable with the TLS acquisition, supported by the higher level and resolution of the camera shooting.
Figure 9. Quality of processed morpho-metric TLS and UAV data on the internal surfaces of the masonry structure. The point cloud from the UAV acquisition
shows a worse resolution in the discrete definition of the main profiles, geometries, and surfaces, largely due to the corner and shadow shooting.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 93
in the UAV photogrammetric processing environment, and the
normal information was corrupted within the different matrixes
of conversion. Following various attempts at comparisons on the
multiple modelling platforms suitable for the managing of
architectural scale data, an additional phase of the re-calculation
of the normal vertex in the UAV point cloud was adopted and
performed directly in the modelling environment, so as to
correctly orient the mesh triangulation and to avoid anomalies of
flipped polygon regions that are not suitable for the mesh
management as a whole. Due to the high density of points and
their unstructured distribution, further complicated by the
geometric articulation of the complex, the automatic normal
calculation tools did not return uniform and coherent results,
fragmenting the orientation of the points and, consequently, the
features of the triangulated mesh. As such, only 70 % of the
points in the UAV-based point cloud (1.426.256 points) were
automatically processed for the normal feature.
To resolve this issue, a single-step triangulation process was
performed with the TLS and UAV integrated point cloud such
that the normal vertex data defined in the TLS point cloud could
mediate the orientation of the entire mesh surface. The
topological meshing procedure was performed using an HD
adaptive triangulation set for the following parameters:
- Geometry capture accuracy: set over 75 % of the details for
shape accuracy.
- Scanner accuracy: this related to the overall accuracy
guaranteed by the source point cloud. Given the source
TLS and UAV data, and the quality/noise of the
integrated aligned database, the accuracy was set to
within 3 mm.
Filtering and selection of source data for the integration procedure.
The second aspect related to the different point spacing
interval of the individual point clouds (on average, 1-3 mm from
the TLS, and 5-10 mm from the UAV) and its influence on the
triangulation of an overall polygonal mesh, where a uniform
reading of structural irregularities was ensured. The two
discontinuous datasets were refined with an overall filtering
performed on the point clouds and were checked via a ‘watertight
remesh’ process on the final mesh, assuming an average edge
length of 5 mm to define a final model compatible with the main
standards of deformation analysis in the area of structural
diagnosis. This range did not compromise the relative
dimensions of the structural damage quantifiable as a whole and
related to the geometry of the construction materials (mainly
bricks and metal chains of larger specific dimensions) and to the
tolerance threshold set for the analysis (5 mm, also compatible
with the standards for elastic safety assessments), Figure 11.
The finalisation of the integrated 3D survey database related
to the Pokcha complex defined a virtual system of the preserved
shape, directing the attention toward the metric-spatial
correspondence of information obtained from the TLS database
and the UAV photogrammetry, calibrated at different reliability
values of space reconstruction from the features of the
instruments applied in the acquisition phase. Specifically, a
morphological reference and registration was developed in terms
of the scale of each structural unit of the built complex. Here, for
the pavilion vault, the two types of data were aligned in terms of
perimetrical boundaries and façades while considering the
deviation accuracy of the discreate surfaces and target control
points. Following this, a segmentation of the overlapped point
cloud was performed, deleting the overlapped areas and
maintaining the quality of the TLS data on the intrados surfaces
and both the TLS and UAV data on the extrados surfaces.
Finally, the set of points was subjected to standard filtering
procedures, both automatic and manual, which were aimed at
cleaning all the spatial information not pertaining to the envelope
of the structural surface, specifically constituted by the infiltrated
vegetation both at the base and at the top of the ruins. Further
attention was paid to the window openings, which were manually
cleaned for a complete reconstruction of the wall thickness. The
triangulation phase of the final integrated database highlighted
certain pieces of missing morphological information, which was
due to the building masonry areas being covered by vegetation
during the survey campaign (removed in the point cloud during
the filtering process). These parts were subsequently integrated
via a fitting of mesh holes according to the geometric primitives
derived from the mesh model. The mesh model obtained by
triangulating and remeshing the surfaces verified the expected
target of morphological detail, conserving both the main
structural profiles and the specific geometry qualities of the
masonry elements, both on the surface and on the edges along
the collapsed region.
Figure 10. Alignment and integration of multi-instrumental point clouds for
the static unit of the central pavilion vault.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 94
5.2. Non-uniform rational basis spline comparison and deviation
analysis
The investigation into the structural instabilities was focused
on the vaulted unit in the central span, with the aim of comparing
the specific imperfect geometry of the vault (affected by
mechanical deformations) with a reliable reconstruction of its
original shape from the in-site dimensions of the ruins. Using the
final mesh model, various section planes were set with an
established pitch of 30 cm along all the surfaces of the unit (vault
and vertical walls), extracting 3D sketch profiles that adhered to
the specific morphological surface and importing them into a
NURBS modelling environment. Using extrusion techniques, the
structural volumes of the pavilion vault were geometrically
reconstructed according to the regular project profiles and
imported into the reverse modelling environment to allow for a
comparison with the irregular mesh surface, Figure 12.
The comparison between the NURBS model and the mesh
model of the vault was performed using the query tool of
deviation available on the platform to compute the discrepancy
between the models’ surfaces. The deviation comparison (made
possible by the same coordinate system maintained between the
models) indicated an overall failure phenomenon that
characterised the entire system of the resistant unit, possibly
attributable to the accidental impact that caused its collapse.
Within it, various specific portions affected by instability could
be observed since their colour map obtained via the deviation
computing was characterised by a colour scale ranging from
green to red, highlighting the attendant on-course kinematic
deformation that is increasing over time. These areas are located
both on the vault and on the supporting walls, within
deformations that are on a scale of centimetres and that are
affecting the main loaded parts of the vaulted system, Figure 13.
Specifically, much like with the collapsed part, the remaining
structure of the vault exhibits a discharging arch not in line with
the overall static configuration. In fact, this shape reached 40 mm
of deformation along entire height, it is subdivided into three
plastic hinges and presents a serious risk of collapse. The main
bearing walls also exhibit a buckling deformation that reached a
maximum of 70 mm, thus increasing the instability of the
structure in the basement area.
As such, the reliable mapping of evolving structural
instabilities using direct survey and mesh models is aimed at
determining the intervention practices involving targeted
consolidation actions for the safety and conservation of ruined
structures.
6. CONCLUSIONS
The modelling performed in terms of the Church of the
Annunciation in Pokcha followed an integrated approach
involving morphometric data from different instrumental point
clouds. An overall mesh triangulation strategy was devised to
produce a reality-based model capable of preserving the
structural irregularity through the mediation of numerical
polygonal surfaces. Specific methodological considerations were
developed for the mesh triangulation of the integrated TLS and
UAV sparse databases. In order to perform an HD mesh
construction, a correct alignment of the points was required as
well as UAV point cloud processing to support the optimisation
of the poly-face orientation in the mesh. Other filtering processes
were implemented in terms of the point cloud, specifically
regarding the presence of opening grids and extensive vegetation,
Figure 11. Final reality-based mesh models following the triangulation and
optimisation processing of the HD mesh. The mesh preserved the specific
details of the structural shape of the structural unit.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 95
to better expose the surface of the structural domain affected by
the decay and the natural elements of the ruin site.
The need for a formal approach to the re-drawing analysis and
intervention for endangered historical sites was the motivation
behind the operational experimentation of morphological-
structural representation in terms of two specific research targets:
1) The planning of a documentary strategy aimed at acquiring
the totality and particularity of the architectural details in
terms of all typological variants (masonry, metal parts, wall
coverings) and collocations (main environments,
underground, in elevation, coverage levels).
2) The convenience of transferring these detailed systems into
suitable morpho-metric products capable of providing
information and analytical opportunities in relation to
historical masonries through graphic representations.
These objectives determined the preference for a 3D
approach to the documentation and visualisation of structures
directly from the data obtained via digital surveys. The shape
representation is preserved in both qualitative and quantitative
structural assessments, aware of the interactions that historical
architecture can exhibit between its individual preserved
components and, in terms of restoration, with the intervention
design, Figure 14.
Due to its complex nature in terms of compatibility and
format size, the type of morphometric data represented by high-
poly mesh models is currently affected by the continuous
developments in the field of Big Data management in relation to
3D databases pertaining to architectural systems and urban
aggregates. This possibility for reverse modelling in terms of
historical architectural heritage allows for revising the
documentation protocols in a field of application of
representative systems for structural diagnostics, to aspire to
technical-professional adoption in the interface with mesh-to-
Figure 12. NURBS modelling of the pavilion vault from the sections extracted
by the mesh surface. The geometric vault was referenced to the mesh model
and was imported in the same workspace.
Figure 13. Deviation computing on the shape features between the models of the pavilion vault, from the HD mesh and from the NURB modelling. The colour
map defines a quantitative reading of the deformed portions from the static configuration, highlighting the kinematic mechanisms that are affecting the vault.
From the comparison, it was possible to identify the areas of instability and to derive diagnostic considerations in relation to the overall static schemes of the
architectural complex as a whole.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 96
FEA and mesh-to-BIM protocols. Despite the national risk, the
lines applied in the digital documentation demonstrate an
attempt to adapt to the provisions of the Digital Agenda 2020 of
the European Commission (also included in the Agenda Culture
2030), regarding the digitation of cultural heritage initiatives for
the planning of intervention, restoration, and conservation. With
the possibility of experimenting virtually within a simulated but
reliable space, the methods and effects of a subsequent physical
implementation of the intervention for heritage buildings
emphasises the opportunities for digital preservation and
eGovernance, expanding the ideas of innovation in relation to
the type and quality of structural survey products in terms of the
curation of digital assets and advanced digitisation. The action of
'structural' documentation, including the knowledge of stress
behaviours and the prediction of damage mechanisms in
buildings, can thus be identified within the European guidelines
for heritage at risk by insisting on the characteristics of
robustness.
ACKNOWLEDGEMENTS
The editorial authorship is granted to Sandro Parrinello for
sections 1, 3 and the conclusions, and to Raffaella De Marco for
sections 2, 4 and 5.
The documentation of the Upper Kama Region is part of a
wider program of activities carried out, since 2013, by DAda-
LAB - University of Pavia (coordinator prof. S. Parrinello) and
the Perm National Research Polytechnic University (coordinator
prof. S. Maximova). The digital survey campaigns of 2018 and
2019 were conducted by Parrinello S., Picchio F., De Marco R.,
Dell’Amico A. Part of the architectural survey documentation
data was processed as part of the activities related to the course
entitled ‘Architectural survey & restoration’ (prof. S. Parrinello,
prof. G. Minutoli) in the Double Degree Italian/Chinese course
in Building Engineering and Architecture at the University of
Pavia.
The surveying, documentation and modelling of
Blagoveshchenskaya Church are part of the EU project,
PROMETHEUS. This project is funded by the EU program
Horizon 2020-R&I-RISE-Research & Innovation Staff
Exchange Marie Skłodowska-Curie and involves the
collaboration between three universities (University of Pavia,
Italy, Polytechnic University of Valencia, Spain, Perm National
Research Polytechnic University Perm National Polytechnic
University Research, Russia) and two companies (EBIME, Spain,
SISMA, Italy).
This project has received funding from the European Union’s
Horizon 2020 research and innovation programme under Marie
Skłodowska-Curie grant agreement No 821870.
REFERENCES
[1] R. de Koster (eds), Guidelines on Cultural Heritage. Technical
tools for heritage conservation and management, JP - EU, 2012,
Online [Accessed 26 March 2021]
https://rm.coe.int/16806ae4a9
[2] United Nations Educational, Scientific and Cultural Organisation,
Operational Guidelines for the Implementation of the World
Heritage Convention, WHC.17/01, UNESCO World Heritage
Centre, Paris, France, 2017.
[3] UNESCO, Recueil de décisions importantes sur la conservation
des biens du patrimoine culturel inscrits sur la Liste du patrimoine
mondial en péril de lUNESCO, WHC-09/33.COM/9, Paris, 2009.
[4] K. Rao, A new paradigm for the identification, nomination and
inscription of properties on the World Heritage List, International
Journal of Heritage Studies 16 (2010), pp. 161-172.
DOI: 10.1080/13527251003620594
[5] L. Toniolo, M. Boriani, G. Guidi (eds), Built Heritage: Monitoring
Conservation Management, Springer, Cham., Switzerland, 2015,
ISBN 978-3-319-08533-3. DOI: 10.1007/978-3-319-08533-3
[6] E. Psychogiopoulou, Cultural Heritage and the EU: Legal
Competences, Instrumental Policies, and the Search for a
European Dimension, in: A. Jakubowski, K. Hausler, F. Fiorentini
(editors), Cultural Heritage in the European Union. A Critical
Inquiry into Law and Policy, Brill Nijhoff, Netherlands, 2019,
ISBN: 978-90-04-36534-6, pp. 57-78.
DOI: 10.1163/9789004365346_012
[7] S. Parrinello, R. De Marco, From the city to the stone: digital
survey for the establishment of structural behaviours in historical
architecture, in: R. Salerno, Drawing as (in)Tangible
Representation, Gangemi Editore, Roma, Italy, 2018, ISBN 978-
88-492-3651-4, pp. 747-754.
[8] B. M. Feilden, J. Jokilehto, Management Guidelines for World
Cultural Heritage Sites, ICCROM, Rome, Italy, 1998. ISBN: 92-
9077-150-X
[9] M. Bruneau, S. E. Chang, R. T. Eguchi, G. C. Lee, T. D. ORourke,
A. M. Reinhorn, M. Shinozuka, K. Tierney, W. A. Wallace, D. von
Figure 14. Properties of the final HD mesh model, which is suitable for compatible computing analysis in FEA workflows.
ACTA IMEKO | www.imeko.org March 2021 | Volume 10 | Number 1 | 97
Winterfeldt, A framework to quantitatively assess and enhance the
seismic resilience of communities, Earthquake Spectra 19(4)
(2003), pp. 733-752.
DOI: http://dx.doi.org/10.1193/1.1623497
[10] A. Giuffrè, C.F. Carocci, Statics and dynamics of historical
masonry buildings, in: proceedings International Workshop on
Structural Restoration of Historical Buildings in Old City Centers,
Heraklion, Crete, 1994, pp. 71-152.
[11] A. M. DAltri, G. Milani, S. de Miranda, G. Castellazzi, V. Sarhosis,
Stability analysis of leaning historic masonry structures,
Automation in Construction Volume 92 (2018), pp. 199-213.
DOI: 10.1016/j.autcon.2018.04.003
[12] F. Parisi, N. Augenti, Earthquake damages to cultural heritage
constructions and simplified assessment of artworks, Engineering
Failure Analysis 34 (2013), pp. 735–760.
DOI: 10.1016/j.engfailanal.2013.01.005
[13] S. Parrinello, F. Picchio, P. Becherini, R. De Marco, The drawn
landscape in 3d databases: the management of complexity and
representation in the historical city, 7th Annual International
Conference on urban Studies & Planning, Atiners Conference
Paper Series, Athens, 2018, pp. 3-26.
DOI: 10.30958/aja.4-3-2
[14] S. Parrinello, F. Picchio, R. De Marco, A. DellAmico,
Documenting the Cultural Heritage Routes. The creation of
informative models of historical Russian churches in the Upper
Kama Region. The International Archives of the
Photogrammetry, Remote Sensing and Spatial Information
Sciences XLII-2/W15 (2019), pp. 887-894.
DOI: 10.5194/isprs-archives-XLII-2-W15-887-2019
[15] W. C. Brumfield, Landmarks of Russian Architecture, Gordon e
Breach, Amsterdam, 1997.
[16] E. N. Shumilov, Orthodox Churches and Temples of Common
Faith in Perm Krai: short historical guide, PKUB, Perm, 2003.
[17] L. A. Shatrov, Monuments of architecture, history, art of Cherdyn
and Cherdyn Region, V.2. Cherdyn, 1997.
[18] A. Guarnieri, F. Pirotti, M. Pontin, A. Vettore, Combined 3D
surveying techniques for structural analysis applications, Proc. of
International Symposium on Photogrammetry and Remote
Sensing (ISPRS), 2005, vol. XXXVI-5/W1, pp. 22-24.
[19] S. Parrinello, F. Picchio, R. De Marco, New generations of digital
databases for the development of architectural urban risk
management, in: C. Gambardella, Le Vie dei Mercanti XV Forum
Internazionale World Heritage and Disaster, Fabbrica della
Conoscenza, vol. 71, 2017, ISBN: 978-88-6542-582-4, pp. 1-10.
Online [Accessed 26 March 2021]
http://www.leviedeimercanti.it/proceedings-xv-forum
[20] M. Tsakiri, D. Lichti, N. Pfeifer, Terrestrial laser scanning for
deformation monitoring, in: Proceedings of 3rd IAG/12th FIG
Symp., Institute of Geodesy and Geophysics, Baden, Germany,
2006, pp.1-10. Online [Accessed 26 March 2021]
https://www.fig.net/resources/proceedings/2006/baden_2006_
comm6/PDF/LS2/Tsakiri.pdf
[21] A. Guarnieri, N. Milan, A. Vettore, Monitoring of complex
structure for structural control using terrestrial laser scanning
(TLS) and photogrammetry, International Journal of Architectural
Heritage 7(1) (2013), pp. 54-67.
DOI: 10.1080/15583058.2011.606595
[22] S. Bertocci, G. Minutoli, G. Pancani, 3D survey and instability
analysis of Romena parish. Disegnare CON 8(14) (2015), pp. 1-20
Online [Accessed 26 March 2021]
http://disegnarecon.univaq.it/ojs/index.php/disegnarecon/artic
le/view/36/31
[23] R. De Marco, Shapes and models: the survey for the study of
structures in historical buildings, in: K. Williams, M. G.
Bevilacqua. Nexus 2018 Architecture and Mathematics -
Conference Book. Kim Williams Books, Pisa, Italy, 2018, ISBN
978-88-88479-47-7, pp. 289-292.
[24] L. De Luca, P. Veron, M. Florenzano, Reverse-engineering of
architectural buildings based on an hybrid modeling approach,
Computers & Graphics 30(2) (2006), pp. 160–176. Online
[Accessed 26 March 2021]
https://hal.archives-ouvertes.fr/hal-01021897
[25] G. Castellazzi, A. M. DAltri, S. de Miranda, F. Ubertini, An
innovative numerical modeling strategy for the structural analysis
of historical monumental buildings, Engineering Structures 132
(2017), pp. 229-248.
DOI: 10.1016/j.engstruct.2016.11.032
[26] M. Centofanti, S. Brusaporci, Architectural 3D modeling in
historical buildings knowledge and restoration processes, in: C.
Gambardella, Le Vie dei Mercanti, X Forum Internazionale di
Studi, More or Less. Fabbrica della Conoscenza, vol. 16, 2012,
ISBN 978-88-6542-129-1, pp. 331-340. Online [Accessed 26
March 2021]
http://www.leviedeimercanti.it/proceeding-x-forum/
[27] S. Parrinello, R. De Marco. From survey to the model: the graphic
transposition of an earthquake, Disegnare Idee Immagini 57
(2018), pp. 70-81.
