Innovative Methodology for the 3D Reconstruction of Body

Geometries using Open-source Software

Javier Tuesta-Guzmán

, William Solórzano-Requejo

1,2 b

, Gustavo Grosso-Salazar

Carlos Ojeda

and Andrés Díaz Lantada

Department of Mechanical and Electrical Engineering, Universidad de Piura, Piura, Peru

ETSI Industriales, Universidad Politécnica de Madrid, Madrid, Spain

carlos.ojeda@udep.edu.pe, andres.diaz@upm.es

Keywords: 3D Scan, Reverse Engineering, Meshroom, Open-source Software, Photogrammetry.

Abstract: Bioengineering teaching has limitations in developing countries due to the inaccessibility of expensive

technology like scanners and commercial software, which holds back progress in the biomedical area because

of a lack of resources. In this work, a new methodology is presented with the aim of obtaining a 3D model of

body part by open-source software: Meshroom

, Meshmixer

, Ultimaker Cura

; and a cell phone camera.

The procedure is based on three methods which were tested: images taken for a short time, burst mode and

video-to-frames. Through the process of reverse engineering photogrammetry, an arm and a foot were

obtained from images and for comparing the model with the real body part, 3D printing was used. The

outstanding method is video-to-frames thanks to the high quality of the generated models and the shortest

reconstruction time it presents. The technique developed can promote the education of engineers in the

biomedical area, also providing an advance for developers with low economic resources, allowing them to

have a new possibility of research.

1 INTRODUCTION

Medical Reverse Engineering (MRE) is highly used

to treat fracture problems, obtaining hurt body parts

in 3D format as STL (Stereo Lithography) or OBJ

(Object) file (Bhatti et al., 2018). MRE is also

fundamental in any medical device personalization

strategy (Ahluwalia et al., 2022). Computer-aided

design software tend to be expensive, which may

limit equal accessibility to personalized healthcare

technologies. Fortunately, different open-source

hardware and software solutions are being developed,

which synergize with the "maker's movement" and

promote equity in all fields of product development.

These technologies allow getting the affected zone,

necessary for the design and manufacture of custom

medical devices. The usage of open-source software

is a proposal for the education framework of

https://orcid.org/0000-0002-6923-7263

https://orcid.org/0000-0002-2989-9166

https://orcid.org/0000-0002-7570-5609

https://orcid.org/0000-0001-6163-5382

https://orcid.org/0000-0002-0358-9186

engineering students due to remote education as a

consequence of the COVID 19 pandemic, which

forced to change the techniques used to teach

(Pokhrel & Chhetri, 2021). Any researchers already

focused their attention on the development of

technologies that could assist students in anatomy

remote education (Iwanaga et al., 2021) as Qlone

for

3D scanning.

Scanning in the biomedical area is a method to

obtain an external body part like legs, arms, hands,

face; necessary to design an external fixation for

closed fracture (Alqahtani et al., 2021) or help relieve

articular injuries (Munoz-Guijosa et al., 2020) but for

this, designers need a special camera to scan it, and

this is a limitation for the developed of personalized

medical technology due to the high cost (Le et al.,

2010). Fortunately, Meshroom

(Griwodz et al.,

2021) and COLMAP

(Schonberger & Frahm, 2016)

162

Tuesta-Guzmán, J., Solórzano-Requejo, W., Grosso-Salazar, G., Ojeda, C. and Díaz Lantada, A.

Innovative Methodology for the 3D Reconstruction of Body Geometries using Open-source Software.

DOI: 10.5220/0010870200003123

In Proceedings of the 15th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2022) - Volume 1: BIODEVICES, pages 162-169

ISBN: 978-989-758-552-4; ISSN: 2184-4305

exist, whose principal function is to do the object

reconstruction from images. This technique is called

photogrammetry and is less common for 3D

reconstruction but a MRE useful tool to get the base

for the biodevices design (Struck et al., 2019). The

photogrammetry for 3D surface scanning allows good

visualization of the objects and people without a high

knowledge about photography can do a

reconstruction (Grabherr et al., 2016).

This project explores innovative approaches to

promote MRE through open-source software. A

procedure relying on the combination of varied

software, including: VLC Media Player 3.0.16

Vetinari

, Meshroom 2021.1.0

, Meshmixer

3.5.474

, and UltimakerCura 4.8.0

, is presented. To

the author’s knowledge, this combination is

innovative and reported for the first time. It is

illustrated through a set of case studies.

The goal is to promote collaborative work,

inspired in projects as UBORA (De Maria et al.,

2020), wanting to reinforce the techniques used for

remote study and have an important impact in

developing countries for people with few resources,

focused on the formation of students, engineers and

doctors who want to research in the biomedical field.

This work is organized as follows: first, materials

and methods used to obtain the arm and foot models

will be explained; secondly, the results and discussion

of the quality of models, and finally the conclusions

of this project.

2 MATERIALS AND METHODS

Figure 1 shows the roadmap to obtain the 3D model,

as a base for designing personalized medical devices,

from images of the body part employing open-source

software. This research assesses the influence of the

method to get the images, in the mesh given by the

programs.

Figure 1: Roadmap.

External devices need a geometrical body

obtained through photogrammetry, which needs

images around the body part, then they are imported

to Meshroom

which reconstructs the scene and

generates an OBJ file, different tests will be done to

define which is the best method together with the

correct number of images. Subsequently,

Meshmixer

was used to optimize, crop, and repair

the 3D model that finally is prepared and printed

through Ultimaker Cura

2.1 Reverse Engineering

2.1.1 Photogrammetry

Photogrammetry can be understood as a methodology

to determine distance relations between different

bodies in space from images. One image provides

two-dimensional coordinates of each point of the

photo, and if it is complemented by a second image

from a different position, so a three-dimensional

coordinate could be calculated for each point (Linder,

2009). It is a powerful tool for reverse engineering to

obtain models of inanimate objects, also could be

used to reconstruct body parts with the correct steps.

Meshroom

is an open-source software for 3D

reconstructions based on the AliceVision framework.

The software works in a base of photogrammetry, and

to make a reconstruction is necessary to configure

nodes that Meshroom

offers. The environment

nodes by default receive the inputs (images),

separating them into groups and highlighting the most

notable features, then generate a points cloud of the

scene in 3D, filter the inconsistencies and finally

generate a mesh and project the textures in the model

(Dong et al., 2021). Meshroom

will be used for the

experiments of bodies part reconstruction adding a

new node to establish the actual scale.

Their minimum requirements are RAM: 8GB,

GPU: NVIDIA CUDA, CPU: Intel or AMD and

works in any operating system.

To take the pictures, it is necessary to use a camera

or multiple cameras (Peyer et al., 2015) which allows

obtaining a detailed view of the body reconstructed.

The cell phone used was a Samsung A50 which has a

triple camera. The primary camera has a resolution of

24.9MP with an aperture of f/1.7. The second (ultra-

wide) and third (depth sensor) ones have an aperture

of f/2.0 with 8MP and 5MP respectively.

2.1.2 Meshroom and Animated Objects

Through Meshroom

, different models can be

obtained of all types of inanimate things as vehicles

(Matys et al., 2021), ship structures (Shah et al.,

2021), and more if given it images of the object from

360° degrees as shown in Figure 2. Meshroom

uses

these images to establish measure relations between

the background and the main piece. For this reason,

while more images are given, the detail of the final

model should be better, but not all is quantity because

Innovative Methodology for the 3D Reconstruction of Body Geometries using Open-source Software

163

if blurred images are used, the final model will be

affected. The difficulty is obtaining a 3D high-quality

model of a self-moving object.

Figure 2: Images surrounds the body.

For example, arm, face, foot, hand, and functional

body parts are considered as self-moving objects, if

someone tries to take images surrounding them, it is

a fact that exists minimal involuntary movements

because of the tired of maintaining these parts in a

special position, necessary to do the images for 2

minutes or more. If these images are given to

Meshroom

, the final model will have all the

background: walls, floor, and other objects minus the

body part. An idea could be to eliminate the

background and only have the body part in the

pictures; the result will be an amorphous model

because distance relations are eliminated, which are

essential for Meshroom

works. Therefore, it is

necessary that the interaction with the participants be

the minimum to avoid these unwanted movements

(Peyer et al., 2015).

Figure 3 shows the Meshroom

interface; the

images generate a point cloud that is exported in OBJ

format. The default model scale is not the real

magnitude. For this reason, it is necessary to use the

CCTAG3 templates (CCTag /MarkersToPrint at

Develop · Alicevision/CCTag, 2021) that were

provided by AliceVision, since Meshroom

needs to

set up the real size.

Figure 3: Meshroom

interface.

2.2 Experiments

Six experiments have been performed with different

methods and several images to generate the best

model as possible. The first method consisted in

taking pictures surrounding the body; the

reconstructions were made with 100 images taken for

4 minutes approximately, then, this experiment was

repeated with half of the images (50) that were taken

from the first 2 minutes. As some involuntary

movements may exist, another approach to take the

pictures should be sought, for this purpose, in the

second method the use of the burst mode of the phone

was performed. Finally, the third method (video-to-

frames), proposed by the authors, is based on

recording a 40-second video, since Meshroom

does

not use videos for reconstruction it must be converted

into frames. To maintain the conditions of each

reconstruction, 100 and 50 images were used for each

method.

To convert the video on frames, VLC Media

Player

was used (Figure 4). This program generates

a high number of images according to the recording

ratio, for example, 100 frames in 30-second video.

Thus, the number of images for the recognition is

maximized and the involuntary movements are

minimized.

BIODEVICES 2022 - 15th International Conference on Biomedical Electronics and Devices

164

Figure 4: Conversion of video-to-frames through VLC

Media Player

3 RESULTS AND DISCUSSION

3.1 Arm

To take the images, the person was sitting down with

their arm supported on a table where the CCTAG3

templates stay collocated with 30cm of distance and

should be visible.

The models were exported from Meshroom® and

imported in Meshmixer® to visualize it, for

understanding the behavior of the models, a similar

sight of the reconstruction will be compared.

Figure 5: Arm reconstruction with (A) 100 and (B) 50

photos, (C) 100 and (D) 50 burst photos, (E) 100 and (F) 50

frames.

Figure 5 shows the different models obtained

through the experiments. In Figure 5A is visible that

the arm was not reconstructed but at least obtain a part

of the neck, shoulder, and a few of the inanimate

objects. In Figure 5B, despite using half of the photos

on purpose to reduce the time to take images (4 to 2

minutes), the result was not good, this model does not

show the arm too. In Figure 5C appears a form similar

to the arm with poorly defined parts (errors in

reconstruction), but the shape is noticeable. In Figure

5D, the use of half of the images did a few differences

defining better some parts of the body but not enough.

Figure 5E presents the arm reconstructed with a void

in the last part of the forearm and Figure 5F present

the arm with nom-uniform areas. Anyway, the

difference with the first four models is noticeable.

In the models from method of photos is notorious

that the presence of the involuntary movements is the

principal problem, for this reason, images should not

be taken for a long time or else not be obtained the

desired model, despite this, it is a necessity a large

number of images for the reconstruction. Burst mode

of phone method presents better models than only

taking photos due to reduction of the capture time, but

not enough to minimize the effect of the involuntary

movements. Also, any images from burst mode could

have low quality when trying surround the main piece

and as a consequence the final models are poorly

detailed and useless. For video-to-frames method was

obtained a good reconstruction, first, the model is

recognizable, there is no doubt that the reconstruction

is an arm; second, the arm was reconstructed

completely and with high quality compared with the

previous methods, third, the surroundings objects

were reconstructed correctly too. It is visible that

these models present non-uniform areas due to in

frames exist a problem of brightness, the recording

place has angles where the arm looks darker, and

deformations occur right at these angles. In other

words, when existing different illumination, the

model will be affected directly, so the lightning must

be constant to preserve the quality of reconstruction.

It visualized that the use of half images has fewer

errors in the result obtained in each technique, for

example, 50 bursts model is better than 100 bursts

model; this may be due to if use an excessive number

of images, results in oversizing whose inconsistencies

are not suppressed in totally in filter nodes.

Table 1 summarizes the results for each

experiment through Meshroom

. It is easily noticed

that the best model was obtained with the method

proposed “video-to-frames”, also the number of

triangles in Table 1 has a relation with the number of

Innovative Methodology for the 3D Reconstruction of Body Geometries using Open-source Software

165

images used, more photos make more reconstruction

of the main piece and its environment.

Table 1: Vertices, triangles, and reconstruction time for

each arm model.

Model Vertices Triangles Time

100 photos 308672 612916 60m 7s

100 bursts 428034 849489 86m 50s

100 frames 138060 274218 36m 56s

50 photos 232459 461799 28m 8s

50 bursts 366186 727762 45m 57s

50 frames 37695 74290 10m 11s

Models with 50 pictures have fewer triangles than

the model with their respective method with 100. It is

not possible to compare the number of triangles of the

six models only having the arm, cropping the OBJ

file, because in some the arm is not completely

reconstructed.

The fastest reconstruction for 50 and 100 images

was the video-to-frames models, so this method

performs a better reconstruction in less time.

3.2 Foot

The experiments were repeated with the foot to

analyze the performance with another body region,

for comparative

purposes. In this case, to took

images, the person stood up and placed their leg near

to CCTAG3 templates which establishes the real

scale.

Figure 6 shows the results of the reconstruction of

each technique and the respective number of images.

Figure 6A presents an amorphous and incomplete

mesh, same problem in Figure 6B but adding that due

to the fewer images, the reconstruction presents a gap

between the foot and calf. For Figure 6C, the calf zone

is visible with protuberances, and it is the same for

the model of Figure 6D but with a gap in the leg. In

Figure 6E, the reconstruction is similar to the 100

burst model but more complete and better defined.

Finally, Figure 6F presents a good quality model but

with more protuberances in the calf zone. Same as the

models of the arm, the best model was generated by

the video-to-frames method.

The results are shown in Table 2. The model with

the fewest triangles and the best reconstruction time

was from video-to-frames. The number of triangles

has a direct impact on the reconstruction time. As this

is minimal, less computational resources are required,

which allows the use of this technique and technology

for educational purposes.

Figure 6: Foot reconstruction with (A) 100 and (B) 50

photos, (C) 100 and (D) 50 burst photos, (E) 100 and (F) 50

frames.

Table 2: Vertices, triangles, and reconstruction time for

each foot model.

Model Vertices Triangles Time

100 photos 477171 949084 94m 16s

100 bursts 461408 918505 76m 6s

100 frames 190621 379804 38m 36s

50 photos 406215 808415 51m 56s

50 bursts 417762 816884 58m 58s

50 frames 138283 275557 15m 43s

Best models will be selected for arm and foot on

purpose to be optimized with Meshmixer

and printed

with Ultimaker Cura

. Both are the models obtained

with 50 frames due to being better reconstruction,

with minor time and weight file.

Meshroom

delivers unsmoothed models with

some surface defects, for this is necessary to use CAD

BIODEVICES 2022 - 15th International Conference on Biomedical Electronics and Devices

166

software to eliminate this problem. Meshmixer

is an

open-source software useful to optimize OBJ/STL

files and will be used to visualize and fix our models.

To optimize arm and foot, both were cropped and

just stayed with body part of interest (Figure 7). Auto-

repair Meshmixer

tools were used to solve mesh

problems if they exist. After, the models are exported

as STL files.

Figure 7: Cropping and optimization with Meshmixer

for

the (A) arm and (B) foot models.

To obtain the physical model, the 3D printing

technique used was Fused Material Deposition

(FDM). Between the steps of this process was

included the lamination in several layers of the STL

models using the Ultimaker Cura

slicer. The G Code

was exported from Ultimaker

to manufacture the 3D

models employing Ender 3 Pro

printer.

The parameters used for the impression were the

following: PLA Flibox Silver, infill 10% and infill

pattern line, wall line count 3, bottom and top layers

4, speed 50 mm/s, retraction activate, fan speed

100%. Supports were used for the foot model.

Figure 8: Comparison between the printed and real arm.

Figure 9: Comparison between the printed and real foot.

Figures 8 and 9 show two views of the 3D printed

model compared with the scanned body for arm and

foot respectively. Printed models are similar to real

animated objects in shape and scale, the printed arm

has the geometry of the forearm, and the fingers are

unified due to Meshroom recognizing them as one

body. For printed foot, the geometry matches with the

real foot. Both models have good quality and if this

methodology continues to be refined, it can be used

to obtain the base bodies for the design of 3D printing

orthoses (Górski et al., 2020), orthopaedic insoles

(Cendrero et al., 2021; Shalamberidze et al., 2021)

and facial protection masks (Morita et al., 2007).

Other authors (Peyer et al., 2015) constructed a

system with 18 cameras to perform 3D reconstruction

from taking simultaneous photos, minimizing the

interaction time with the body. Their surface mesh

presents high-resolution, but it is not easy for students

from developing countries to obtain these systems.

The images used for these experiments had a

quality of 25 MP and the results were good, if the

number of pixels was increased, the results would

have more detail and a better surface finish.

4 CONCLUSIONS

Printed models of OBJ/STL files generated with the

proposed methodology have high quality for

educational purposes and transmit the essence of the

main piece coinciding with reality. This opens a

world of possibilities for the reconstruction of

animated objects from images as the design of

biodevices or remote education through laboratories

in study centres for the reconstruction of 3D models

without an expensive camera.

Innovative Methodology for the 3D Reconstruction of Body Geometries using Open-source Software

167

The video-to-frames technique proved more

reliable than only photos and burst mode. Their

models show the essence of the shapes in both prints,

also minimize the time for the reconstruction but still

have a few protuberances due to the change of

illumination concerning the angle of the shot and the

limited number of pixels of the images. The method

still could be better if images are taken in other

prepared places where the lighting remains constant

facilitating the reconstruction process and if a

professional camera will be employed to record the

video. Between the data used, exists blurry images

which could be eliminated manually, but exist

algorithms that can do this automatically like Fast

Fourier Transform or Variance of Laplacian which

can help to improve the proposed technique in the

future.

Among main limitations of the study, it is

important to mention that a simple camera has been

used, which may affect precision, although at the

same time its use puts forward the possibility of using

very low-cost hardware and software for promoting

MRE. Regarding future studies, our proposal is to

progress in processes for automated generation of 3D

models, to enhance precision and to employ these and

similar reconstructions, as input for the design of

personalized medical devices, such as splints, insoles,

braces and multiple orthoses.

ACKNOWLEDGEMENTS

The authors would like to thank the support of the

Biomechanics Group of the Universidad de Piura and

to the Product Development Laboratory of the

Universidad Politécnica de Madrid. Also, we

acknowledge the support of reviewers and their

relevant recommendations, which help to do a more

consistent and detailed paper.

REFERENCES

Ahluwalia, A., De Maria, C., & Díaz Lantada, A. (Eds.).

(2022). Engineering Open-Source Medical Devices (1st

ed.). Springer International Publishing.

https://link.springer.com/book/9783030793623

Alqahtani, M. S., Al-Tamimi, A. A., Hassan, M. H., Liu, F.,

& Bartolo, P. (2021). Optimization of a Patient-Specific

External Fixation Device for Lower Limb Injuries.

Polymers, 13(16), 2661. https://doi.org/10.3390/

polym13162661

Bhatti, A., Syed, N. A., & John, P. (2018). Chapter 5—

Reverse Engineering and Its Applications. In D. Barh

& V. Azevedo (Eds.), Omics Technologies and Bio-

Engineering (pp. 95–110). Academic Press.

https://doi.org/10.1016/B978-0-12-804659-3.00005-1

CCTag/markersToPrint at develop · alicevision/CCTag.

(2021). GitHub. https://github.com/alicevision/CCTag

Cendrero, A. M., Fortunato, G. M., Munoz-Guijosa, J. M.,

De Maria, C., & Díaz Lantada, A. (2021). Benefits of

Non-Planar Printing Strategies Towards Eco-Efficient

3D Printing. Sustainability, 13(4), 1599.

https://doi.org/10.3390/su13041599

De Maria, C., Di Pietro, L., Lantada, A. D., Ravizza, A.,

Mridha, M., Torop, J., Madete, J., Makobore, P., &

Ahluwalia, A. (2020). The UBORA E-Infrastructure for

Open Source Innovation in Medical Technology. In J.

Henriques, N. Neves, & P. de Carvalho (Eds.), XV

Mediterranean Conference on Medical and Biological

Engineering and Computing – MEDICON 2019 (pp.

878–882). Springer International Publishing.

https://doi.org/10.1007/978-3-030-31635-8_106

Dong, C., Liang, W., & Xu, P. (2021). 3D refined

reconstruction for brushes based on multiple images.

2021 3rd International Conference on Advances in

Computer Technology, Information Science and

Communication (CTISC), 309–314. https://doi.org/

10.1109/CTISC52352.2021.00063

Górski, F., Wichniarek, R., Kuczko, W., Żukowska, M.,

Lulkiewicz, M., & Zawadzki, P. (2020). Experimental

Studies on 3D Printing of Automatically Designed

Customized Wrist-Hand Orthoses. Materials, 13(18),

4091. https://doi.org/10.3390/ma13184091

Grabherr, S., Baumann, P., Minoiu, C., Fahrni, S., &

Mangin, P. (2016). Post-mortem imaging in forensic

investigations: Current utility, limitations, and ongoing

developments. Research and Reports in Forensic

Medical Science, 6, 25–37. https://doi.org/10.2147/

RRFMS.S93974

Griwodz, C., Gasparini, S., Calvet, L., Gurdjos, P., Castan,

F., Maujean, B., De Lillo, G., & Lanthony, Y. (2021).

AliceVision Meshroom: An open-source 3D

reconstruction pipeline. Proceedings of the 12th ACM

Multimedia Systems Conference, 241–247.

https://doi.org/10.1145/3458305.3478443

Iwanaga, J., Terada, S., Kim, H.-J., Tabira, Y., Arakawa,

T., Watanabe, K., Dumont, A. S., & Tubbs, R. S.

(2021). Easy three-dimensional scanning technology

for anatomy education using a free cellphone app.

Clinical Anatomy, 34(6), 910–918. https://doi.org/

10.1002/ca.23753

Le, C., Jos, V. S., Le, T. H., Lam, K., Soe, S., Zlatov, N.,

Le, T. P., & Pham, D. T. (2010). Medical reverse

engineering applications and methods [Conference

Proceedings]. University of Greenwich; INCDMTM.

https://gala.gre.ac.uk/id/eprint/11735/

Linder, W. (2009). Introduction. In W. Linder (Ed.), Digital

Photogrammetry: A Practical Course (pp. 1–17).

Springer. https://doi.org/10.1007/978-3-540-92725-9_1

Matys, M., Krajcovic, M., & Gabajova, G. (2021). Creating

3D models of transportation vehicles using

photogrammetry. Transportation Research Procedia,

55, 584–591. https://doi.org/10.1016/j.trpro.2021.07.0

BIODEVICES 2022 - 15th International Conference on Biomedical Electronics and Devices

168

Morita, R., Shimada, K., & Kawakami, S. (2007). Facial

Protection Masks After Fracture Treatment of the Nasal

Bone to Prevent Re-injury in Contact Sports. Journal of

Craniofacial Surgery, 18(1), 143–145. https://doi.org/

10.1097/01.scs.0000246729.23483.87

Munoz-Guijosa, J. M., Zapata Martínez, R., Martínez

Cendrero, A., & Díaz Lantada, A. (2020). Rapid

Prototyping of Personalized Articular Orthoses by

Lamination of Composite Fibers upon 3D-Printed

Molds. Materials, 13(4), 939. https://doi.org/

10.3390/ma13040939

Peyer, K. E., Morris, M., & Sellers, W. I. (2015). Subject-

specific body segment parameter estimation using 3D

photogrammetry with multiple cameras. PeerJ, 3, e831.

https://doi.org/10.7717/peerj.831

Pokhrel, S., & Chhetri, R. (2021). A Literature Review on

Impact of COVID-19 Pandemic on Teaching and

Learning. Higher Education for the Future, 8(1), 133–

141. https://doi.org/10.1177/2347631120983481

Schonberger, J. L., & Frahm, J.-M. (2016). Structure-from-

Motion Revisited. 2016 IEEE Conference on Computer

Vision and Pattern Recognition (CVPR), 4104–4113.

https://doi.org/10.1109/CVPR.2016.445

Shah, F. M., Gaggero, T., Gaiotti, M., & Rizzo, C. M.

(2021). Condition assessment of ship structure using

robot assisted 3D-reconstruction. Ship Technology

Research, 0(0), 1–18. https://doi.org/10.1080/09377

255.2021.1872219

Shalamberidze, M., Sokhadze, Z., & Tatvidze, M. (2021).

The Design of Individual Orthopedic Insoles for the

Patients with Diabetic Foot Using Integral Curves to

Describe the Plantar Over-Pressure Areas.

Computational and Mathematical Methods in

Medicine, 2021, e9061241. https://doi.org/10.1155/

2021/9061241

Struck, R., Cordoni, S., Aliotta, S., Pérez-Pachón, L., &

Gröning, F. (2019). Application of Photogrammetry in

Biomedical Science. In P. M. Rea (Ed.), Biomedical

Visualisation: Volume 1 (pp. 121–130). Springer

International Publishing. https://doi.org/10.1007/978-

3-030-06070-1_10

Innovative Methodology for the 3D Reconstruction of Body Geometries using Open-source Software

169