Content uploaded by Ivan Dunđer
Author content
All content in this area was uploaded by Ivan Dunđer on Nov 15, 2022
Content may be subject to copyright.
TEM Journal. Volume 8, Issue 1, Pages 94-101, ISSN 2217-8309, DOI: 10.18421/TEM81-13, February 2019.
94 TEM Journal – Volume 8 / Number 1 / 2019.
Application of Photogrammetry in 3D
Scanning of Physical Objects
Ivan Reljić 1, Ivan Dunđer 1
1 Department of Information and Communication Sciences, Faculty of Humanities and Social Sciences,
University of Zagreb, Ivana Lučića 3, 10000 Zagreb, Croatia
Abstract – This paper explains the 3D scanning
procedure of creating a virtual 3D model from
photographs by using a process called
photogrammetry. It starts by giving a technical
explanation of different technologies for 3D scanning,
explains why photogrammetry was chosen and gives
general specifications of hardware and software used
in the process. The whole procedure is then thoroughly
shown step by step on a physical object, and in the end
an analysis of the generated 3D model and its
variations is given.
Keywords – 3D scanning, photogrammetry, 3D
modelling, virtual 3D models, information technology
1. Introduction
3D scanning is rapidly rising in popularity all over
the world because it is becoming more accessible
and, maybe even more importantly, easier to use for
people who are not specialists. Since the cost of the
equipment is dropping while the quality is increasing,
it is becoming a technology of choice for
documenting in many different areas, such as cultural
heritage, archaeology, civil engineering, medicine,
multimedia etc.
DOI: 10.18421/TEM81-13
https://dx.doi.org/10.18421/TEM81-13
Corresponding author: Ivan Reljić,
Faculty of Humanities and Social Sciences, University of
Zagreb, Zagreb, Croatia
Email: ireljic@gmail.com
Received: 11 December 2018.
Accepted: 19 January 2019.
Published: 27 February 2019.
© 2019 Ivan Reljić, Ivan Dunđer; published by
UIKTEN. This work is licensed under the Creative
Commons Attribution-NonCommercial-NoDerivs 3.0
License.
The article is published with Open Access
at www.temjournal.com
This research aims to explain the process of
creating a 3D scan from photographs in a concise
way along with some of the issues that might arise
and how to solve them.
This paper derives from a project aiming to
connect Croatian culture, heritage and tourism
through a process of digitisation. The first step of this
project includes choosing an adequate physical object
and digitising it.
The authors of this paper chose to digitise the bust
of Ivan Meštrović. The physical object is located in
Zagreb, Croatia and the author is Stipe Sikirica. It
was installed in 1984.
Ivan Meštrović was chosen as he is one of the
greatest Croatian artists, author of many works in
well-known locations, such as:
- “Zdenac života” (eng. The Well of Life) from
1905, located in front of the Croatian
National Theatre in Zagreb, Croatia;
- relief “Seljaci” (eng. Peasants) from 1907,
which is overlooking the Ban Josip Jelačić
square, the central square in Zagreb, Croatia;
- statue “Grgur Ninski” from 1929, in Split,
Croatia;
- sculpture “Job” from 1945, which is located
in Syracuse, New York, United States of
America.
It is appropriate for a sculpture in his image to be
carried over to a new visual medium, this time
virtual. This is done as part of an ongoing project
which aims to create an integrated process by which
Croatian culture, artwork and heritage locations can
be promoted, all by using new technologies.
The intention of the initial phase of the project is to
create a virtual 3D model by utilising easily
accessible technology (personal computers, mobile
phone cameras etc.), and assessing the effectiveness
of such an approach.
2. Related work
One research presents different uses for
photogrammetry for the purpose of conservation and
study of heritage [1]. Three case studies with various
TEM Journal. Volume 8, Issue 1, Pages 94-101, ISSN 2217-8309, DOI: 10.18421/TEM81-13, February 2019.
TEM Journal – Volume 8 / Number 1 / 2019. 95
purposes were used, their processes explained and
results compared to each other.
Another research describes two approaches to
creating a 3D model from photographs [2] for the
purpose of cultural heritage documentation and
analysis. The approaches are tested on two case
studies and they are compared by using RMSE (Root
Mean Square Error), which shows the difference of
length between two points on photographs and the
length of the same points on the 3D model.
An algorithm for reconstructing 3D structure from
unordered photographs is presented in [3]. The
theory of Structure-from-Motion (SfM) and its
implementation is explained in detail along with a
software implementation of the whole procedure
named COLMAP.
One study highlights the potential of multi-image
photogrammetry as a rapid survey tool and for
community engagement [4]. It shows that the
technology can be used with very limited budget both
for the survey part and for the post-processing of the
data while also maintaining a short time frame.
A qualitative and quantitative assessment has
shown that low-cost photogrammetry can be used for
documentation and preservation of historical and
cultural heritage with a case study of a World War II
fortification [5].
A qualitative and quantitative comparison between
3D scanning approaches has been made using a laser
scanner and photogrammetry, in which a reference
3D model is created using a 3D scanner and then
compared to a 3D model created from photographs of
the object [6].
According to [7], a methodology to collect 3D data
of objects with complex geometry through a case
study of a church was presented. The proposed
process clearly shows that using readily available
equipment can match expensive specialised
equipment, and, in some cases, even outperform it,
e.g. when reaching inaccessible areas.
3. Different approaches to creating a 3D model
Different approaches to creating a 3D model out of
physical objects are sorted by accuracy and presented
hereafter.
3.1. Manual modelling
Refers to 3D modelling by hand in one of 3D
modelling software packages [8]. If a physical object
is being modelled, usually photographic references of
the object are being used in the process. This
approach is frequently used for objects that have
been physically damaged and when there are not
enough photographs for the automatic 3D
reconstruction, such as in the case of Bamiyan
Buddhas in Afghanistan, which were destroyed by
dynamite in 2001 [9].
3.2. 3D scanning
There are a lot of technologies but the ones most
used are contact 3D scanning, non-contact active and
non-contact passive 3D imaging solutions [10]:
Contact: for scanning of small and simple objects
(e.g. coins). The scan is created by using a thin probe
which touches the physical object in many
predefined positions. The resulting 3D model is a set
of virtual points in the same positions where the
probe touched the object.
Non-contact active: devices emit light and
measure the time needed to get back to the device,
and in that way, they reconstruct the distance from
the scanner for all the points on the object. This
group consists of laser scanners, LiDAR (Light
Detection And Ranging) and structured light
scanners. Such devices are quite expensive but
precise. Their downside is that, when scanning larger
objects that cannot be rotated on a turntable (like e.g.
monuments), the devices have to be carried to
different locations around the object which is time-
consuming and physically exerting. In cramped
spaces sometimes even that is not possible.
Non-contact passive: devices that receive light
(such as photo and video cameras) reconstruct a 3D
object by using the photographs made from different
angles around that object. This approach is the
simplest to use, but it depends on multiple factors:
quality of their sensor, lighting conditions, a large
number of photographs is required, operator
experience, software tools that can convert
photographs to 3D models etc. Such devices are very
mobile so they can be used in different conditions
like scanning from the air or for underwater
scanning. This approach also includes
photogrammetry [11]. Photogrammetry is the science
of obtaining reliable information about the properties
of surfaces and objects without physical contact with
the objects, and of measuring and interpreting this
information [12].
Oftentimes it is possible to find specific 3D models
on the internet. Such models have been modelled
either by hand or through 3D scanning. Available
models on the internet, that are similar to a new 3D
modelling task, can sometimes be adjusted to fit the
requirements of a particular project in order to save
time.
TEM Journal. Volume 8, Issue 1, Pages 94-101, ISSN 2217-8309, DOI: 10.18421/TEM81-13, February 2019.
96 TEM Journal – Volume 8 / Number 1 / 2019.
4. Research
This research will be discussed in detail in the
following subsections.
4.1. Technology employed
For the purposes of this research the
photogrammetry method of 3D scanning was used.
Such a procedure is simple, does not require
specialised technical equipment and the fieldwork,
i.e. photographing the object, can be done in a
reasonable amount of time by an experienced
operator.
4.2. Technical data
In this experiment a mobile phone camera with the
following technical specifications was used for
photographing the bust of Ivan Meštrović in Zagreb,
Croatia: 16 MP resolution, 2.2 aperture, focal length
31 mm, sensor size 1/2.6”, pixel size 1,12 µm.
The total number of photographs taken was 199.
As for the lighting conditions, the object was in
shadow during sunset.
The statue is composed of a bronze bust on a stone
pedestal. The whole size of the statue is
approximately 108 cm W x 203 cm H x 80 cm L.
The bust itself is 82 cm W x 61 cm H x 52 cm L.
For the reconstruction process of creating a 3D
model from photographs, a specialised proprietary
software was used called Agisoft PhotoScan (now
Agisoft Metashape): “Agisoft PhotoScan is a stand-
alone software product that performs
photogrammetric processing of digital images and
generates 3D spatial data to be used in GIS
applications, cultural heritage documentation, and
visual effects production, as well as for indirect
measurements of objects of various scales.” [13].
The computer used for the reconstruction has the
following specifications: 4,2 GHz quad core
processor, 16 GB of DDR4 RAM, graphics card with
4 GB of VRAM.
When it comes to location, the bust of Ivan
Meštrović is in a place which has no physical
obstructions in the vicinity, which means a 3D model
viewable from all sides could be created. Also, it was
close to the operator so any kind of potential changes
and reworks were not a problem.
As for the physical characteristics, the total height
of the statue is such that the operator could access the
whole surface without using any kind of technical
tools (e.g. tripod, monopod, ladder and such), which
significantly decreased the time required for this
phase of the project.
In terms of appearance, the surface of the selected
bust is of low visual frequency, i.e. smooth, so there
was a lesser probability of visual artefacts (error in
the calculation and display of the surface). Although
the monotony of a surface makes the reconstruction
harder, as the 3D modelling software has problems
differentiating points on the surface [14], in this
particular case, there was enough diversity to
successfully perform the reconstruction.
In the case of lighting conditions, the photographs
were created in November 2018, in the afternoon,
when there was no direct light which creates
shadows, but there was still enough ambient light to
acquire photographs of sufficient quality. The whole
procedure of photographing was done during the
weekend, because it is less probable that a random
passer-by might appear on some of the photographs.
Any kind of change between photographs
interferes with the matching algorithm which looks
for similarities between photographs, so the
appearance of a person, even in one of the
photographs, negatively affects the algorithm. If that
person appears on more than one photograph while
moving around the object (i.e. appears to be in
multiple locations on different photographs), it is
possible that the process of 3D model reconstruction
fails completely. In that case, photographs can be
retouched in one of the specialised software
applications to remove the person, but that is a time-
consuming process and it still does not guarantee
success. In this research, since the time of
photographing was carefully chosen and the operator
was paying attention to the surrounding area, there
were no people appearing in the photographs.
Regarding the photograph selection and data set
compilation, after the process of taking photographs,
it is necessary to go through all of the photographs by
hand, and to remove all the problematic ones which
might negatively impact the reconstruction process.
Such procedure requires an experienced operator in
order to minimise issues that might arise in the later
part of the process. Not only should the operator
know how to take good photographs and inspect
them visually, but the same person should have a
good knowledge of the whole reconstruction process
in order to recognise visual cues that could impact
the quality of the final 3D model.
4.3. Process and methods of photographing
The most important element of the
photogrammetric process is to get as much coverage
of the object as possible. It is not enough to get the
whole object on a minimum number of photographs
– a lot of attention must be put into getting a lot of
overlap between photographs [15].
The algorithm that reconstructs the form of the
physical object compares photographs of the object
and puts them into pairs, so a bigger overlap allows
TEM Journal. Volume 8, Issue 1, Pages 94-101, ISSN 2217-8309, DOI: 10.18421/TEM81-13, February 2019.
TEM Journal – Volume 8 / Number 1 / 2019. 97
the algorithm to compute a larger number of shared
points from which it creates a model later on [15].
A larger overlap also means that every point on the
object is visible from multiple photographs, so if one
point is visible on at least three photographs, that
point can be precisely triangulated in three-
dimensional space [15]. Such photographs were
taken for this research as well.
Another important element is the camera’s
resolution as the number of pixels has a direct impact
on the number of points the algorithm can analyse,
and, by that, it affects the final resolution of the 3D
model [15]. The camera used in this research has 16
MP and, as such, has more than sufficient resolution
for a texture that satisfies the needs of this
experiment.
Alongside having the resolution impact the
model’s point density, the quality of the texture is
also a result of the same photographs [16].
Certainly, the number of pixels on a photograph is
not necessarily a measure of a photograph’s quality.
It is better to have pixels with good focus, i.e. a
quality lens and sensor of the camera that create the
least amount of noise [17]. It is better to have less
pixels of high quality than a large number of low-
quality pixels. A large number of low-quality pixels
implicates longer processing time to recreate a 3D
model that will, eventually, have worse surface
quality and a bad texture [17].
Besides the coverage from photographs and their
resolution, lighting of the object is also of great
importance [18]. As the complexity of colours makes
it easier to find shared points, so do the shadows on
the object help with reconstructing the form of the
object. Having too many shadows is bad, because
they have a negative influence on the reconstruction
of both form and texture [19].
In controlled conditions there are ways to
completely remove shadows while also having a
good reconstruction of the 3D model and texture.
That is achieved by a combination of well-placed
lights, a ring flash and using a polarisation filter
which filters reflections from the object [20]. In
outdoor conditions it is optimal to take photographs
in the middle of a cloudy summer day, because there
is a lot of ambient light colours and details are
showing up on the object without having shadows.
For a simple reconstruction process, it is preferable
to have the whole object on all of the photographs,
although this is not necessary. In this way the
algorithm easily recognises all the locations of points
on the object, and can use the silhouette of the object
to further simplify the process.
The best overlap of photographs is when they have
a radial offset of 10 to 30 degrees, while the
maximum offset is 5 to 45 degrees [15]. When the
offset is larger than 30 degrees, there is not enough
overlap and the 3D model will be reconstructed with
holes (missing parts) or it might not be reconstructed
at all [15]. With offset of less than 10 degrees, there
is a lot of redundancy between photographs which
greatly increases the required time for reconstruction
without actually increasing the quality of the 3D
model [15].
The position of the camera is easy to control if
there is always an equal distance from the object and
the camera is always at the same height level. After
one circle around the object, the height at which the
camera is held changes and another circle of
photographs is made around the object.
It is important to tilt the camera towards the
previous circle of photographs so they would all be
radially offset vertically as well. If the circular sets of
photographs are always facing the horizon
(cylindrical camera positions), it is possible to have
bad overlap. With photographs that have both
horizontal and vertical offset, the overlap will be the
best, and so the 3D model will be of a lot higher
quality.
In summary, this research satisfied all of the
conditions:
- the photographs had good focus;
- noise present in the photographs was in
low amounts due to favourable lighting;
- there was no discrepancy of the surface
colour of the physical object due to
absence of cast shadow;
- the object had low amounts of cast
shadows which improved the texture
quality;
- more than sufficient overlap between
photographs was present;
- camera positioning was hemispherical in
regards to the centre of the object.
5. Selection of photographs and preparation of
the 3D modelling data set
When using a camera with autofocus, it is common
that some of the photographs come out blurred. Such
photographs do not contribute to the process of
reconstructing the 3D model, because the exact
location of points on the model cannot be extracted
from these photographs.
Furthermore, at the same time such photographs
affect the quality of the texture as well. Although
blurred photographs do not help with reconstruction
of the model, if put into the same reconstruction data
set with other photographs, they will also be included
in the reconstruction process.
Therefore, it is important that even before the
reconstruction process starts, the operator goes
through all the photographs and leaves out the ones
that are blurred.
TEM Journal. Volume 8, Issue 1, Pages 94-101, ISSN 2217-8309, DOI: 10.18421/TEM81-13, February 2019.
98 TEM Journal – Volume 8 / Number 1 / 2019.
With partially blurred photographs, the problem
gets bigger the closer the blur is to the centre of the
photograph. In underwater photogrammetry, all the
photographs have blurring on the outer sides due to
the properties of light passing underwater, but it is
possible to create a 3D model if the central area of all
the photographs is of adequate quality [21].
When the operator identifies a bad photograph,
there are several steps he can take:
- remove the photograph from the
reconstruction data set;
- make a new photograph of the object from
the same angle as the original photograph;
- leave the photograph in the data set and try
to reconstruct a 3D model despite the
possibility that a part of the model does not
get reconstructed.
All of these approaches have some negative
consequences for the whole project. Since this
research had favourable conditions in regards to the
location of the object, complexity of the objects
form, ambient lighting, obstructions, operator
experience etc., there were no such photographs that
would require any of the aforementioned procedures.
Removing the photograph from the reconstruction
data set usually means that this particular part of the
surface is not covered on at least three photographs.
That can be avoided by creating enough
redundancy during the process of taking photographs
of the object. When creating redundancy for only one
photograph, if the maximum radial offset between
two photographs is 30 degrees, that means that every
photo should have a maximum of 15 degrees offset
from the previous photograph. If there is a need for
redundancy of two photographs in a row, the
maximum offset would be 10 degrees. In this
research, due to a relatively low number of
photographs required, it was simple to add more
photographs so the radial offset was around 8
degrees.
Whenever possible, it is best to create a new
photograph instead of leaving the blurred one in the
data set, but for different reasons that is not possible
in many situations:
- the object is physically far away from the
operator so retaking the photographs creates
unforeseen costs in time and finances;
- a lot of heritage objects, especially ones
being scanned due to danger of deterioration,
require complicated procedures to obtain a
permit for accessing the artefacts, and even
then there is a limited timeframe in which
access is granted, so if time runs out, the
whole procedure has to be repeated anew;
- quality equipment is often being rented for a
specific project in order to cut expenses, and
it would have to be rented again just for a
few photographs;
- weather conditions may change in the
meantime: object might be covered in snow,
there is a long rain period, weather changes
from cloudy to sunny and vice versa, so new
photographs do not match the old ones etc.;
- the object gets damaged or even completely
destroyed, whether by natural occurrences or
intentionally;
- politics and law might change making the
physical object inaccessible.
In cases when blurring of photographs is of lesser
intensity or on certain parts of object, one could try
to leave the photograph in the data set and run the
reconstruction with such inferior data, because even a
lesser surface quality of the 3D model and texture
looks better to the end user than completely missing
parts of the 3D model.
This research had no blurred photographs so the
issues mentioned were avoided.
Holes in the model should always be avoided if
possible, as they break the illusion of realism of the
3D model.
A completely described 3D model looks like it is
solid on the inside, but a model with holes shows the
user that it is actually empty, and therefore a person
will find it hard to enjoy the view. Moreover, the
person might start to think about the technology
instead of the content.
If the operator wishes to cap the holes, he would
have to recreate the surface by hand and then project
that part of the texture from surrounding
photographs. This is quite labour-intensive and at the
same time imprecise, so for some purposes that
model becomes unusable. Holes in the 3D model are
sometimes impossible to avoid due to the nature of
the physical objects, in which case the models are not
shown from those sides, but just the visible ones.
Those types of objects include: statues and busts on a
fixed base, buildings, carved rock, geographical
locations etc. The 3D model created during this
research had no holes in it except for the one on the
bottom (the ground on which the statue stands) and,
as such, it could not have been avoided.
Besides autofocus, cameras oftentimes have
automatic exposition which can lead to colour
discrepancy. In regards to the light source, one part
of the model is more exposed to light than the other.
When changing the light source is not an option, like
it is with sunlight, the camera’s white balance should
be set up. If the camera does not have that option,
some of the photographs will be light and the others
dark. The camera used in this research does not have
TEM Journal. Volume 8, Issue 1, Pages 94-101, ISSN 2217-8309, DOI: 10.18421/TEM81-13, February 2019.
TEM Journal – Volume 8 / Number 1 / 2019. 99
auto white balance (AWB), therefore there were
some discrepancies between photographs. When
photographs of the same object have different
exposure values, exposition should be manually
changed afterwards. All the photographs were loaded
into a software for editing photographs where all the
expositions were normalised using a semi-automatic
procedure: a middle value of exposition was
extracted and the photographs with noticeable
expositions were then adjusted to that middle value.
Differing lighting creates problems for the
reconstruction algorithm, because it uses shadows as
a reference for generating surface detail of the object,
which means that differently coloured shadows on
the object can lead to imprecise reconstruction of
form. Because there was enough ambient light and
the exposure was manually normalised, the shadows
were even and, as such, did not create any issues
during the reconstruction phase.
A 3D model reconstructed from differently lit
photographs, apart from surface irregularities, will
also have differently coloured texture. Such 3D
model will then look like it has a part of its surface in
shadow. This was avoided in this research by
carefully choosing the time of day when to take
photographs and manually correcting some of them
in the software for editing photographs before
starting the reconstruction.
If the virtual lighting of the 3D model, i.e. light
source vector, does not match real lighting on the
photographs, the 3D model will look like it has
shadows in different directions, and that is something
a user, unfamiliar with how this technology works,
cannot understand, and the 3D model will look
completely unnatural (even if he is not able to
explain why) [15]. Since the object in this research
was photographed with no direct sunlight on it and
the photographs were manually edited in a software
for editing photographs, the 3D model did not have
visible shadows in the texture so it can be used with
any kind of virtual lighting.
5.1. Analysis of the object selection process
When using photogrammetry, all the photographs
which meet the following criteria should be
discarded:
- photographs which are out of focus:
photographs should have the same depth of
field;
- photographs with too much exposition: this
“burns” the photograph and there is a loss of
colour information;
- photographs which do not receive enough
light: it is hard to differentiate what is
shadow and what is the object;
- photographs that are redundant: photographs
made from almost the exact angle, as that
increases processing time without improving
the quality of the 3D model.
Since this research used a relatively simple subject
and had an experienced operator, there were no
discarded photographs.
6. Adjustment of photographs
As far as the photograph adjustment process,
photographs can sometimes have smaller
imperfections that can be removed in any of the
photo processing applications – lesser blur, noise that
appears in low-light conditions, unbalanced
exposition, rotation of the camera etc. But, in this
research, the only thing that needed to be adjusted
was the exposition difference between photographs,
which was expected due to equipment used and the
lighting in which the object was photographed.
The exposition was normalised across photographs
in such a manner that overexposed photographs were
slightly darkened, and underexposed photographs
were slightly brightened, which, in the end, turned
out to be beneficial to the quality of the texture of the
3D model.
7. Analysis of the 3D model creation process and
discussion of the results
When it comes to the 3D model reconstruction, the
whole data set consisting of 199 photographs was
loaded into the selected specialised photogrammetry
software application for 3D modelling.
Surface of the 3D model had a quality that can be
used for different purposes, especially for online
display which was one of the main purposes of the
project. The quality of the texture was higher than
needed but, since the images the texture was made
from do not take up a lot of space for storage, they
will be kept for future reference.
Qualitative analysis has shown that the described
approach can provide data of higher quality than
needed even when using non-specialised equipment
for data acquisition and processing.
A model was reconstructed in different resolutions
of the 3D model, and in different resolutions of the
texture. All the variations of the same model are
shown in Table 1.
The best result was achieved when all the
photographs from this experiment were used and the
3D model was reconstructed in the highest quality
with highest quality texture (size increase of ca.
242%), but in regards to the needs of the research
versus processing time for the highest quality, that
was not necessary.
TEM Journal. Volume 8, Issue 1, Pages 94-101, ISSN 2217-8309, DOI: 10.18421/TEM81-13, February 2019.
100 TEM Journal – Volume 8 / Number 1 / 2019.
Table 1. Variations of the reconstructed 3D model.
Model
quality
Model
file
size
Texture
size
Total
Increase
Medium +
4k texture 36,9
MB 2,92
MB 39,82
MB Base
model
Medium +
8k texture 36,9
MB 8,47
MB 45,37
MB 13,94%
Medium +
16k texture 36,9
MB 24,1
MB 61,00
MB 53,19%
High + 4k
texture 112
MB 2,92
MB 114,92
MB 188,60%
High + 8k
texture 112
MB 8,47
MB 120,47
MB 202,54%
High + 16k
texture 112
MB 24,1
MB 136,1
MB 241,79%
The best ratio of reconstruction time and quality is
when the whole set of photographs is used and the
3D model is reconstructed in medium quality with a
texture in medium quality (size increase of ca. 14%).
Such model requires less processing time and loads
faster so it is easier to browse and edit collections of
multiple models. They can also be uploaded faster to
online services for displaying 3D models. For the
same reason, they load faster on slower devices, such
as mobile phones and tablets.
In this research, as the most suitable 3D model, the
one with the best ratio of reconstruction time and
quality was chosen (Medium + 8k texture). It has the
following specifications:
- 3D model composed of 274.911 points
(548.114 polygons);
- chosen file format: .OBJ;
- 3D model file size: 45,37 MB;
- texture with a resolution of 8K (8.000 x
8.000 pixels);
- single texture file in .JPG format;
- texture file size: 8,47 MB.
For the model and texture file, the formats that were
chosen are .OBJ and .JPG because they are:
- industry standards;
- most widespread in the 3D scanning
community and for general use;
- formats for storage, editing, viewing and
displaying 3D models;
- cross-platform;
- for the stated reasons most suited for this
project.
8. Conclusion
This research has shown a complete process of
creating a virtual 3D model, from approaching the
physical object to the final virtual 3D model. The
assumption was successfully confirmed that even
when using limited equipment (e.g. a mobile phone
camera) a quality model can be created that is
valuable for further use.
The obtained 3D model is split into information
about the three-dimensional form of the object
(virtual record of all the points in space) and
information about the colour of the object (its
texture).
Due to development of software tools and
computer hardware, photogrammetry as a 3D
scanning approach is becoming an increasingly
useful tool in many areas in which it was not present
until not long ago.
Through this research it was shown that many
technologies that were restrictive until recently can
be used now, with a greater focus on bringing
cultural heritage to a larger audience.
This may, therefore, warrant future research on
numerous aspects of 3D model generation and
application. It should be investigated to what extent a
3D model without texture could be used in further
analysis of the physical form of the object: e.g. for
virtual simulations of other materials, for the
reconstruction of damaged parts, for the simulation
of deterioration and outside influence, for adding a
different texture etc.
Moreover, more extensive comparative analyses of
various 3D modelling software should be carried out
in the future, in order to identify the necessary
software prerequisites to create a 3D model with
included texture that could be used for online display
as a singular object or as a part of a larger virtual
space alongside similar object (e.g. a virtual
museum).
Acknowledgements
This paper is part of the project “Application of
technology in promoting sustainable development of
tourism and culture (techCULTOUR)”, which is trying to
connect Croatian culture, heritage and tourism with the
help of modern 3D technologies. The project relies on
digitisation processes and attempts to present the results
of digitisation to a wide interested audience and to all
possible stakeholders. The project is supported by the
University of Zagreb, Croatia.
TEM Journal. Volume 8, Issue 1, Pages 94-101, ISSN 2217-8309, DOI: 10.18421/TEM81-13, February 2019.
TEM Journal – Volume 8 / Number 1 / 2019. 101
References
[1]. Bazán, I. R., & Vita, G. E. E. (2017). The utility of the
application of new three-dimensional technologies for
the study and dissemination of heritage from a
historical-technical perspective: case studies. Ge-
conservación, 1(11), 208-213.
[2]. Aicardi, I., Chiabrando, F., Lingua, A. M., & Noardo,
F. (2018). Recent trends in cultural heritage 3D
survey: The photogrammetric computer vision
approach. Journal of Cultural Heritage, 32, 257-266.
[3]. Schonberger, J. L., & Frahm, J. M. (2016). Structure-
from-motion revisited. In Proceedings of the IEEE
Conference on Computer Vision and Pattern
Recognition (pp. 4104-4113).
[4]. McCarthy, J. (2014). Multi-image photogrammetry as
a practical tool for cultural heritage survey and
community engagement. Journal of Archaeological
Science, 43, 175-185.
[5]. Shults, R., Krelshtein, P., Kravchenko, I., Rogoza, O.
& Kyselov, O. (2017). Low-cost Photogrammetry for
Culture Heritage. “Environmental Engineering” 10th
International Conference.
[6]. González-Merino, R., Fraile, A. D., Pérez, J. A., &
Sánchez-López, E. M. (2017). Validation of
photogrammetry techniques performed on two lead
ingots assigned to Linares Historical
Heritage. Procedia Manufacturing, 13, 1405-1412.
[7]. Robleda Prieto, G., & Pérez Ramos, A. (2015).
Modeling And Accuracy Assessment for 3d-Virtual
Reconstruction in Cultural Heritage using Low-Cost
Photogrammetry: Surveying of the “Santa María
Azogue” Church’s Front. 3d-Arch 2015 “3D virtual
reconstruction and visualization of complex
architecture, 263-270.
[8]. Zara, J. (2004, May). Virtual reality and cultural
heritage on the web. In Proceedings of the7th
International Conference on Computer Graphics and
Artificial Intelligence., 101-112.
[9]. Grün, A., Remondino, F., & Zhang, L. (2002).
Reconstruction of the great Buddha of Bamiyan,
Afghanistan. International Archives of
Photogrammetry, Remote Sensing and Spatial
Information Sciences, 34, 363-368.
[10]. Curless, B. (1999). From range scans to 3D
models. ACM SIGGRAPH Computer Graphics, 33(4),
38-41.
[11]. El-Hakim, S. F., Beraldin, J. A., & Blais, F. (1995,
December). Comparative evaluation of the
performance of passive and active 3D vision systems.
In Digital Photogrammetry and Remote
Sensing'95 (Vol. 2646, pp. 14-26). International
Society for Optics and Photonics.
[12]. Schenk, T. (2005). Introduction to
photogrammetry. The Ohio State University,
Columbus, 106.
[13]. Metashape, URL: http://www.agisoft.com/
[14]. James D. W., Eckermann, J., Belblidia, F. & Sienz, J.
(2015). Point cloud data from Photogrammetry
techniques to generate 3D Geometry. Proceedings of
the 23rd UK Conference of the Association for
Computational Mechanics in Engineering.
[15]. Li, J., Li, E., Chen, Y. & Xu, L. (2010). Visual 3D
Modeling from Images and Videos. Intel Labs China.
[16]. Jennings, A., & Black, J. (2012). Texture-based
photogrammetry accuracy on curved surfaces. AIAA
journal, 50(5), 1060-1071.
[17]. Gasparovic, M., & Gajski, D. (2016). Testing of
Image Quality Parameters of Digital Cameras for
Photogrammetric Surveying with Unmanned
Aircrafts. Geodetski List, 70(3), 253-266.
[18]. Westoby, M. J, Brasington, J., Glasser, N. F.,
Hambrey, M. J. & Reynolds, J. M. (2012). “Structure-
from-Motion” photogrammetry: A low-cost, effective
tool for geoscience applications. Geomorphology,
179, 300-314.
[19]. Li, Y., Gong P. & Sasagawa, T (2005). Integrated
shadow removal based on photogrammetry and image
analysis. International Journal of Remote Sensing
26(18), 3911-3929.
[20]. Wells, J., Jones, T., & Danehy, P. (2005, January).
Polarization and color filtering applied to enhance
photogrammetric measurements of reflective surfaces.
In 46th AIAA/ASME/ASCE/AHS/ASC Structures,
Structural Dynamics and Materials Conference (p.
1887).
[21]. Drap, P. (2012). Underwater Photogrammetry for
Archaeology. Special Applications of
Photogrammetry, 114.