See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/321341934

3D Low-cost scanning systems development for mechanical parts

digitalization

Conference Paper · November 2017

CITATIONS READS

0 375

5 authors, including:

Erald Piperi Luigi Maria Galantucci

Universiteti Politeknik i Tiranës Politecnico di Bari
24 PUBLICATIONS   45 CITATIONS    185 PUBLICATIONS   1,207 CITATIONS   

SEE PROFILE SEE PROFILE

Ilo Bodi Tatjana Spahiu

Universiteti Politeknik i Tiranës Polytechnic University of Tirana
8 PUBLICATIONS   42 CITATIONS    18 PUBLICATIONS   34 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

An Attempt for Developing Albanian Anthropometric System within a Pilot Project View project

Creation of a Reverse Engineering Laboratory View project

All content following this page was uploaded by Tatjana Spahiu on 01 December 2017.

The user has requested enhancement of the downloaded file.

1st International Conference “Engineering and Entrepreneurship” Proceedings ICEE-2017

3D Low-cost scanning systems development for mechanical parts digi-

talization

E. Piperi1, L. M. Galantucci2, J. Kaçani1, I. Bodi1 and T. Spahiu3

1
Polytechnic University of Tirana, Faculty of Mechanical Engineering, Department of Pro-
duction and Management, “Mother Tereza” Square, No. 1, Tirana, Albania
2
Dipartimento di Meccanica, Matematica e Management – Politecnico di Bari, Italy
3
Faculty of Mechanical Engineering, Textile and Fashion Department in Polytechnic Univer-
sity of Tirana

epiperi@fim.edu.al, luigimaria.galantucci@poliba.it

Abstract
Engineers in different areas of production increasingly use 3D scanning techniques, where speeding up the
development process is one of the key factors for being competitive in today’s market. Virtualizations of
physical objects, to recreate parts where drawing or CAD data are missing, or making changes to the exist-
ing one is one of the main uses. From a wide range of commercial scanners used for digitalization, 3D low-
cost scanning systems have gained a high interest.
The aim of this paper is to show the implementation of two low cost scanning systems for creating digital
replicas of mechanical parts. The process of digital data acquisition, processing and creating 3D virtual
models will be presented. A good implementation of light conditions, sensor type/size of digital camera in
photogrammetric system can achieve high Level Of Detail (LOD) compared with structured light system.
The displacement map generated from texture increase the resolution geometry details as accurately as pos-
sible. Also for small parts photogrammetry is more flexible.

Keywords: 3D white light scanning, photogrammetry, reverse engineering.

1 Introduction
3D scanning techniques used for digital data acquisition of different objects are increasingly
used by engineers in different areas of production where speeding up the process of product
development is one of the key factors for being competitive in today’s market. Various appli-
cations can be found in different areas, as: mechanical engineering, automotive industry, cul-
tural heritage documentation, orthodontics, apparel industry, etc. Virtualizations of physical
objects, to recreate parts where drawing or CAD data are missing, or making changes to the
existing one is one of the main uses. Depending from the interaction of sensor and the surface
of the objects, these systems can be divided as contact or non-contact 3D scanners. Another
classification of non-contact 3D scanners is active and passive scanning systems. Selecting
the appropriate scanner based on performance and cost for digital data acquisition it’s not an
easy job. The first digital data are taken as point cloud information and then they are trans-
formed into a mesh. 3D models created after data processing can be used for different appli-
cations, such new product development, and simulation, performing data comparison, or re-
producing the object by rapid prototyping techniques.
Structured Light scanners projects patterns of light onto the objects while digital cameras ac-
quire images of the distorted pattern reflected by the object surface. These systems are based
on triangulation principle for measuring objects shapes. Low-cost systems developed by us-
ing off-the-shelf hardware as webcam, projectors, etc., give additional opportunities to gather
3D information for further use in different applications. They can be used in cultural heritage
documentation [1], [2], [3], [4], [5], human body parts [6], [7], [8], [9], or full body [10],
[11], creating CAD models of mechanical parts, [12], [13], industrial rubber parts [14].
Photogrammetry is another technique for object digitalization: it creates 3D models form 2D
images taken in 360o, with overlapping parts of the images. Cultural heritage documentation

pg. 1
1st International Conference Proceedings ICEE-2017

is one of the main applications of photogrammetric techniques [15], for reproducing artefacts
without complex hardware or software [16].
Another application of photogrammetry is the creation of custom made garments by using 3D
body models [17]. Other authors have presented the development of a photogrammetric tech-
nique which is suitable to capture data even in sub millimetre-sized features [18], [19]. Dif-
ferent parameters effect on the accuracy of photogrammetric measures during the process of
digital data acquisition. The focal length and the zoom of the lens, the resolution of the sen-
sors are important factors that influence the reconstruction results of complex or simple shape
objects [20] or jewlery [21]. Photogrammetric systems are an accepted tool in many areas of
industrial manufacturing [22]. Low cost scanning systems based on close-range photogram-
metry have many advantages respect to laser and structured light scanners [23]. Analysing the
implementation of close range photogrammetry and laser scanning techniques, they showed
similar results in point’s clouds [24] and 3D digital models [25].

2 Methodology
In this paper two low-cost 3D scanning systems have been implemented for mechanical parts
digitalisation. The first is a 3D active lighting system, and is an upgrade of an existing setup
done producing a new steel frame to incorporate two active cameras. The second is a photo-
grammetric system done by using a DSLR camera and appropriate lighting conditions. The
3D data created have been processed for reverse engineering data manipulation.

2.1 Active scanning system

The active scanning system implemented consists on the same hardware and software as is
was presented for footwear digitalization [6]. In this new application the setup give the possi-
bility to add another camera (model uEye UI-1480-C – Resolution: 2580x1920, Sensor Size
½”, Sensor Type: CMOS). For testing the performance of pattern quality, an upgraded projec-
tor with HD resolution (1920x1080) was used, as shown in Figure 1.

Figure 1. Dual camera Structured Light System: a) design and b) implementation of the sys-
tem.

Active scanning systems results to be problematic with reflective object surface. A thinner
layer of white coating spray must be used to avoid problems encountered during the scanning
process of mechanical parts as depicted in Figure 2.
It’s not the purpose of this work to test the accuracy of the system setup but it can be consid-
ered in the range between 0.05 and 0.1mm. These results are very sensitive to board used for
calibration, triangulation setup, lighting conditions etc. A further work will show in details
these influence on 3Dscan.

pg. 2
1st International Conference “Engineering and Entrepreneurship” Proceedings ICEE-2017

Figure 2. a) Non coated original part (upper) and single scan; b) Spray coated part (upper)
and single scan.

The output could be elaborated with different techniques of Reverse Engineering for creating
a precise model of original part. In Figure 3, different steps of data manipulation are shown,
from point cloud to solid model, with optimised topology.

Figure 3. From point cloud to optimised topology.

2.2 Passive photogrammetric scanning system

The proposed methodology is based on the following steps: (i) work piece position, lighting
and camera setup; (ii) RAW data capturing; (iii) photogrammetric elaboration. A DSLR cam-
era NIKON D5300 (24 megapixels, APS-C CMOS sensor) equiped with NIKKOR AF-S 50
mm 1:1.4 G lens type are used with a remote camera control. All the setup is enclosed in a
white box 1x1x1m with halogen lighting lamps inside the platform and with adjustable dis-
tance for cutting off the reflection as much as possible. As shown in Figure 4, a total number
of 113 single images were elaborated on Agisoft Photoscan [26] to create a 3D model of a
differential car block. In Figure 5 are presented close-up view of level of details (LOD) to
show the mesh quality from data manipulation. As can be seen the texture quality generated
with this system is better compared with Structured Light system. Agisoft Photoscan is based

pg. 3
1st International Conference Proceedings ICEE-2017

on the structure-from-motion (SFM) and dense multi-view 3D reconstruction (DMVR) algo-

rithms [25].

Figure 4. 3D reconstruction of a differential car block.

Figure 5. LOD of photogrammetric system.

Another 3D photogrammetry solution which can elaborate a big amount of single images
without the necessity to have a high cost PC is ContextCapture [27]. This is done by tiling
option which does not require higher hardware capabilities. As shown in Figure 6 andFigure
7 the 3D object of a diesel engine with real dimension of 1050x1295x760mm, is created by
487 single images and 158 volumetric tiling blocs.

pg. 4
1st International Conference “Engineering and Entrepreneurship” Proceedings ICEE-2017

Figure 6. 3D model of diesel engine by ContextCapture.

Figure 7. Close-up LOD generated with ContextCapture.

3 Results & Discussion

This work illustrated the implementation of two different scanning systems for digitalization
of mechanical parts. The texture images generated by the reconstruction algorithm of
Agisoft™, can be used as displacement map for transforming the 3D model into a high reso-
lution geometry detail as accurately as possible. The lighting conditions has been an im-
portant factor regarding reflection intensity, resulting in bad photo quality/alignment and
dense cloud interpolations. The RAW data give the possibility to remove the chromatic aber-
ration distortion and to reduce the reflective effect, and to increase photo quality. Photo-
grammetric scanning requires very simple equipment, and allows to turn the camera around
the object to scan the entire work piece, allowing the reconstruction in the 3D model also of
undercuts and blind parts. It is possible to obtain better resolution changing the distance from
the camera to the object, the number of photos and the resolution of the sensor.
For what concerns the structured light system, surface treatment of mechanical parts with
coated spray improves the quality of mesh data for Reverse Engineering process. It can be
noticed that the use of HD projector does not increase 3D data resolution. The most sensitive
parameter is the resolution of the active sensors. By using two active sensors, alignment pro-
cess was more accurate. Figure 8 shows good results of photogrammetric system compared
with structured light. They can be further elaborated if high LOD are needed.

pg. 5
1st International Conference Proceedings ICEE-2017

Figure 8. a) Single scan with Structured Light Scanning System versus b) 3D Photogramme-
try reconstruction.

4 Conclusions
The aim of this paper was to show the implementation of two low cost scanning systems for
creating digital replicas of mechanical parts. The good implementation of light conditions,
sensor type/size of digital camera in photogrammetric system can achieve higher LOD com-
pared with structured light system. The displacement map generated from texture increase the
resolution geometry details as accurately as possible. Photogrammetry is more simple and
flexible, also for small parts. It was noticed that using a HD projector in active scanning sys-
tems does not increase 3D data resolution. The most sensitive was the resolution of active
sensor. By using two active sensors, alignment process was more accurate.
Results about data comparison in terms of accuracy, resolution and other features will show
the advantages and disadvantages of these scanning systems, which are part of another pro-
ject.

5 References

[1] C. Rocchini, P. Cignoni, C. Montani, P. Pingi and R. Scopigno, "A low cost 3D scanner based on
structured light," The Eurographics Association and Blackwell Publishers , vol. 20, no. 3, 2001.
[2] A. Georgopoulos, C. Ioannidis and A. Valanis, "Assesing the performance of a low-cost structured light
scanner," International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences,
vol. XXXVIII, no. 5, 2010.
[3] S. P. Mecherron, T. Gernat and J.-J. Hublin, "Structured light scanning for high-resolution documentation
of in situ archaeological finds," Journal of Archaeological Science , vol. 36, pp. 19-24, 2009.
[4] D. Akca, "3D modeling of cultural heritage objects with a structured light system," Mediterranean
Arhaeology and Archaeometry, vol. 12, no. 1, pp. 139-152, 2012.
[5] I. Kalisperakis, L. Gramatikopoulos, E. Petsa and G. Karras, "A structured light approch for the
reconstruction of complex objects," Geoinformatics FCE CTU, vol. 6, pp. 259-266, 2011.
[6] E. Piperi, L. M. Galantucci, J. Kaçani, E. Shehi and T. Spahiu, "From 3D foot scans to footwear designing
& production," in 6-th INTERNATIONAL CONFERENCE OF TEXTILE, Tirana, ALBANIA, 2014.
[7] I. Stancic, J. Music and V. Zanchi, "Improved structured light 3D scanner with application to
anthropometric parameter estimation," Measurement, vol. 46, pp. 716-726, 2013.
[8] A. Tortschanoff, A. Fritz and R. Leitner, "Compact Low-cost Scanner for 3D-Reconstruction of Body
Parts with Structured Light Illumination," nternational Journal of Bio-Science and Bio-Technology, vol. 6,
no. 5, pp. 13-22, 2014.

pg. 6
 1st International Conference “Engineering and Entrepreneurship” Proceedings ICEE-2017

[9] J. Čarnický and D. Chorvát Jr., "Three-dimensional measurement of human face with structured-light
illumination," Measurement Science Review, vol. 6, no. 2, 2006.
[10] W. Yu and B. Xu, "A Portable Stereo Vision System for Whole Body Surface Imaging," Image Vis
Comput., vol. 28, no. 4, pp. 605-613, 2011.
[11] T. Spahiu, J. Kaҫani, E. Shehi and E. Piperi, "3D body scanning technique for anthropometric
measurements and custom clothing designing," in Developing Third Mission Activities in, Durres, Albania
, 2014.
[12] L. M. Galantucci, E. Piperi, F. Lavecchia and A. Zhavo, "Semi-Automatic Low cost 3D Laser scanning
systems for reverse engineering," in 3rd CIRP Global Web Conference, 2014.
[13] D. Page, A. Koschan, S. Voisin, A. Ngozi and M. Abidi, "3D CAD model generation of mechanical parts
using coded-pattern projection and laser triangulation systems," Assembly Automation, vol. 25, no. 3, pp.
230-238, 2005.
[14] H. Cui, Y. Li and C. Li, "Strategy for 3D Reconstruction of Industrial Rubber Part," JOURNAL OF
COMPUTERS, vol. 7, no. 2, pp. 458-463, 2012.
[15] P. Bourke, "Automatic 3D reconstruction: An exploration of the state of the art," GSTF Journal on
Computing , vol. 2, no. 3, pp. 71-75, 2012.
[16] J. Kaufman, M. Clement and A. E. Rennie, "Reverse Engineering Using Close Range Photogrametry for
Additive Manufactured Reproduction of Egiptian Artifacts and other Objects d'art," Journal of Computing
and Information Science in Engineering , vol. 15, 2015.
[17] G. Percoco, "Digital close range photogrammetry for 3D body scanning for custom-made garments," The
photogrammetric record, vol. 17, 2011.
[18] L. M. Galantucci, M. Pesce and F. Lavechia, "A powerful scanning methodology for 3D measurements of
small parts with complex surfaces and sub millimeter-sized features, based on close range
photogrammetry," Precision Engineering, vol. 43, pp. 211-219, 2016.
[19] L. M. Galantucci, F. Lavecchia and G. Percoco, "Multistack Close Range Photogrammetry for Low Cost
Submillimeter Metrology," Journal of Computing and Information Science in Engineering , vol. 13, 2013.
[20] "Accurancy Issues of Digital Photogrametry for 3D Digitalization of Industrial Products," Revue
Internationale de Ingegnerie Numerique, vol. 2, no. 1-2, pp. 29-40, 2006.
[21] E. Piperi, L. M. Galantucci, J. Kaçani, I. Bodi and T. Spahiu, "From 3D scanning to 3D printing:
Application in fashion industry," in 7th International Conference of Textile, Tirana, ALBANIA, 2016.
[22] W. Bösemann, "Industrial Photogrammetry - Accepted metrology tool or exotic niche," The International
Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vols. XLI-B5, pp. 15-
24, 2016.
[23] M. Pesce, L. M. Galantucci, G. Percoco and F. Lavecchia, "A low-cost multi camera 3D scanning system
for quality measurement of non-static subjects," in 3rd CIRP Global Web Conference, 2015.
[24] H. A. Incekara and Z. D. Seker, "Comparison of point clouds produced from close-range photogrammetry
and laser scanning," in 37th ASIAN CONFERENCE ON REMOTE SENSING, COLOMBO-SRI LANKA,
2016.
[25] T. Segreto, A. Bottillo, R. Teti, M. L. Galantucci, F. Lavecchia and B. M. Galantucci, "Non-contact
reverse engineering modeling for additive manufacturing of down scaled cultural artefacts," Elsevier B.V,
vol. 62, no. 10th CIRP Conference on Intelligent Computation in Manufacturing Engineering - CIRP
ICME '16, p. 481 – 486, 2017.
[26] [Online]. Available: http://www.agisoft.com/. [Accessed September 2017].
[27] [Online]. Available: https://www.bentley.com/en/products/brands/contextcapture. [Accessed September
2017].

pg. 7
View publication stats