301
1 INTRODUCTION
The main feature of maritime, air and land
navigation,aswellassurveyingistheextensiveuse
of 2D mapping imaging [28, 29, 16, 5, 21]. One of
their characteristics is the use of specific signs and
abbreviations[3,18]thatsignificantlylimittherange
of users to professionals in the areas of a
ctivity
requiring the use of maps. At the end of the last
century,thankstothedevelopmentofGNSSsystems,
therewasaverydynamicgrowthinthenumberof
peopleusingunprofessionallyavarietyofmapping
imaging. The increase in accuracy of the location
positioning while using the GPS system up to 13
meters (p
=0.95) [25] has enabled the precise
positioning of vehicles (within a single lane)‐the
effect was the development of systems of mapping
imaging for public navigation [23] most frequently
connectedwithgeolocationandvehiclenavigation
Themainlimitationof2Dmappingimagingisno
possibilit
yofdrawingupamapwithoutdistortions
(angular, distance and area) [7, 10, 13]. It makes
spatialorientationdifficultforunprofessionalpeople
associated with navigation. Therefore it can be
concluded that the ideal navigation map would
represent a 3D visualization of the surrounding
world,fromwhichalltheimperfections of2Dma
ps
havebeenremoved.Perhaps thisisoneofthefactors
that caused at the beginning of the twenty‐first
century intensive work on the development of 3D
imagingforalltypesofnavigation.
Thepaperpresentsandcomparesthemethodsfor
creating 3D models of objects [2] for navigation
applicat
ions. It presents an example of modeling
objects with the use of: technical documentation,
direct measurement with a handheld rangefinder,
photogrammetric methods, and laser scanning of a
two‐storey building measuring 30.8m × 11.6m,
locatedattheUniversityofWarmiaandMazuryin
Olsztyn. The hence developed vector 3D model
includes external and int
ernal part of the building
withoutobjectsontheroof.
Modeling 3D Objects for Navigation Purposes Using
Laser Scanning
C.Specht&P.Dąbrowski
GdyniaMaritimeUniversity,Gdynia,Poland
A.Dumalski
UniversityofWarmiaandMazury,Olsztyn,Poland
K.Hejbudzka
StudioA+GInteriorDesign,InteriorPhotography,Scanning3d,LandSurveying,Olsztyn,Poland
ABSTRACT:Thepaperdiscussesthecreationof3dmodelsandtheirapplicationsinnavigation.Itcontainsa
review of available methods and geometric data sources, focusing mostly on terrestrial laser scanning. It
presentsdetaileddescription,fromfieldsurveytonumericalelab
oration,howtoconstructaccuratemodelofa
typical few storey building as a hypothetical reference in complex building navigation. Hence, the paper
presentsfieldswhere3dmodelsarebeingusedandtheirpotentialnewapplications.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 10
Number 2
June 2016
DOI:10.12716/1001.10.02.12
302
2 MODELING3DOBJECTSWITHPUBLICLY
AVAILABLEMETHODS
Thesimplest,availablevirtuallyforeveryonemethod
of developing 3D models is a short‐range
photogrammetry. Pictures taken from a distance of
several to tens of meters from different angles are
obtainedusinghand‐heldcameras,optionallyplaced
on tripods. A prerequisite
is to provide partial
coverage of adjacent photographs. Characteristic
pointsappearinginbothimagesareadoptedtocarry
out the mutual orientation of the images.
Determinationofthepositionofthecamera at each
photo is not necessary. The three‐dimensional
coordinates are obtained from the constructed
photogrammetricmodel[27].
Themethodofashort‐
range photogrammetry is widely used in high‐
precision industrial measurements, to create three‐
dimensional models and orthogonal projections of
objects. Presented in Fig. 1. the model of the main
buildingofGdyniaMaritimeUniversitywascarried
out only with the use of images posted on the
Internet,withthesupportofimaginingfromGoogle
Earth.
Figure1. 3D Model of the main building of Gdynia
Maritime University developed with the use of
photogrammetrictechniques[source:ownstudy]
An alternative approach was used in the paper
[22],whichshowstwowaystoproduce3Dmodelsof
objects. The first used the direct measurement
(inventory)usingopticalmethods,whilethe second
was based on the existing technical documentation.
Figure2showsamodeloftheFacultyofNavigation
of Gdynia
Maritime University made using a
handheldlaserrangefinder.Themeasurementswere
supplementedwithimagesshotwithacamera–they
helped to acquire the texture of individual parts of
the building as well as some architectural details
which were incorporated into the realized 3D solid
object. It can be assumed
that the completed
inventoryassuredaccuracynoworsethan10cm.
Figure2. Selected architectural details of the Faculty of
NavigationofGdyniaMaritimeUniversitybuilding[22]
Adifferentapproachwasusedwhencreating3D
models of characters of navigation aids on the
approachtoNowyPortinGdansk[22].Inthiscase
the possibility of making direct measurements is
virtuallyimpossible,sotheoptimalmethodistouse
thetechnicaldocumentationofindividualcharacters
of navigational aids,
which was obtained from the
Maritime Office in Gdynia. Marking of the
navigational approach to Nowy Portin Gdanskare
nowoldbuoysPM‐2and PM‐3 and thenewbuoys
with the articulated structure and signs of
entrance/inputheads.Figure3presentsthetechnical
documentation and a 3D
model developed on its
basis.
Figure3. Technical documentation and made on its basis
model of an articulated structure buoy of the approach
routetoNowyPortinGdańsk[22]
3 TERRESTRIALLASERSCANNINGASA
METHODOFOBTAININGGEO‐SPATIAL
DATA
Terrestriallaserscanningisamethodof obtaining spatial
coordinatesofobjects,inwhichthroughthelaserdistance
measurement (laser scanner‐object) and the angular
orientation of the device are achieved ortho‐Cartesian
303
coordinatesofthemeasuredpointsinthelocalcoordinate
system of the instrument. The laser scanner is equipped
withmechanismsforemissionofabeamhorizontallyand
vertically.Thelaserbeamsetsonarotatingatahighspeed
optical element, which by rotation around the horizontal
axiscausesitsreflection
atagivenangle[26].Inthisway
themeasurementofpointslyinginthesameverticalplane
istaken.Next,asecondmechanismrotatestheinstrument
byasmallangularintervalinahorizontalplanearoundthe
vertical axis of the instrument and the measurement is
repeated [1]. This
process is carried out in a defined
angular range including the object of measurement. The
speedofobtaineddataisverylarge[4]andmaybeupto1
millionpointspersecond.
Thescanningtechnologywascreatedinthe60ʹsof
the20thcentury.Theexactscanningoftheobjectwas
a very laborious process due to technological
difficulties.Theintroductionofcomputerprocessing
of results has enabled the creation of accurate and
complexobjectmodels.In1994therewascreatedthe
first accurate and fast scanner to measure small
objects using a linear laser beam. In 1996 they
constructed a manually operated scanner that in
addition to geometric data obtained information
aboutthedegreeofreflectionofthelaserbeamonthe
measured surface [8]. These scanners worked in
desktop stationary
mode. The measurement was
taken in laboratory environment. First scanners
working outdoors, placed on tripods, appeared in
1997. These were massive pulse scanners with
externalpowersupplyanddatastoragemodules.In
2003,thefirstphasescannersarrived.Stilltheyused
externalmeasuringmodules,butmeasurementswere
realized much faster. In
scanners produced since
2007 was initiated the integration of memory
modules and internal power supply. There were
introduced solutions used in conventional
measurement methods. The latest generation of
scanners, which appeared first in 2009, are
instruments with integrated memory and power
supply modules, as well as a connected camera,
which
again improved the functional parameters of
the scanner. Some scanners have the ability to
measure multiple reflection of the same laser beam
[19]whichaffectsaccuracy.
Today,intheterrestriallaserscanning, depending
ontheformofthemeasurementsignal,areusedtwo
types of these devices – pulse scanners (Time
of
Flight‐TOF) and phase scanners (PhaseShift‐PS).
Theresultofmeasurementisacloudofpoints(Fig.
4)constitutingasetofpointswithassociatedspatial
coordinatesandvaluesofintensityofreflection[20].
The number of points obtained during the
measurement of a typical object often reaches
millions
innumbers.
The scan speed of pulse scanners ranges from
5,000 to 125,000 points/sec in the latest models [14,
24]. Pulse scanners are characterized by a greater
maximumscanning range than phase scanners [14].
However, the greater range of pulse scanners is
obtainedattheexpenseofspeedofdataacquisition.
The phase measurement is definitely faster. The
scanning speed of a pulse scanner ranges from
120,000to976,000points/sec[14,24].
Figure4.Exampleofacloudofpointsresultingfromlaser
scanning–afragmentofGSMmast
The most advanced laser scanner solution is its
mobile version designed for gathering data for 3D
modelingoflargeareas.Itsessentialelementsare:the
sensorconsistingofseverallaserheadsoscillatingin
the range of 360 degrees (Fig. 5, left), a GNSS
positioning system based typically onactive
surveying networks
along with the radio line RTK
(GPRS) and the IMU inertial unit. Optionally, the
systemiscomplementedwithanodometerrecording
informationonacarwheel rotation,whichallowsto
increasetheaccuracyofpositioningofthevehicleat
the stage of post‐processing. The mobile scanner is
complemented with
the high resolution (2MP)
panoramiccamerasystemallowing theacquisitionof
imageswithmorethan80%coverageofthesphere.
Thissystem allows fullsynchronization of acquired
images, giving them geo‐referencing and scanning
frequencyof20Hzataspeedofeachoftheheadsof
36,000 points/sec. The inertial
unit (IMU) supports
positioning at the frequency of 200 Hz, ensuring
accuracy of 13‐24 cm at 1 minute loss of tracking
GNSSsignals.Thelatestdevelopmentinthefieldof
laserscanningisahandscanner.Thisisahandytool
able to raise the point cloud at the
speed of 43200
points/secwitharangeofapprox.30m,witha point
accuracyof30mm.Aspecialfeatureofthisscanner
is the ability to implement measurements without
geo‐referencing, which predestines it to create 3D
modelsinsideobjects(Fig.5,right).
Figure5. Mobile (left) and handheld (right) laser scanner
fromTrimble
Alaserscannermeasuringaccuracydependsona
varietyoffactors.Ofkeyimportancehereareangular
304
measurement accuracies: horizontal and vertical, as
well as distances. Another factor is the resolution,
which is a derivative of the laser spot size and the
angularintervalofameasurementbeingmade.This
parameterdeterminestheabilityoftheinstrumentto
detect and measure small objects. The surface from
which
the laser is reflected also has an impact on
measurement accuracy. The color and texture of an
object affect the level of intensity of the reflected
signal.Studiescarriedouthavedemonstratedbetter
parametersofreflectingsignalsbybrightsurfaces[6,
12].Equallyimportant,fromthepointofviewof
the
reflectedsignallevel,isthematerialfromwhichthe
object is made [1, 9]. The accuracy of the
measurement,themeasuringrange,andtherecorded
number of points are also affected by weather
conditions[11].Anotheradditionalfactorinfluencing
thequalityandaccuracyofscanningistheshapeof
the
objectbeing measured. These factors may cause
the point cloud noise, distortion at the edges and
other interference. The test of comprehensive and
standardizedverificationofthesefactorswascarried
out at the Institute for Spatial Information and
Measurement Technology at the University of
Applied Sciences in Mainz [4]. They
constructed a
special research laboratory and analyzed several
models of laserscanners. Similar comparisons were
presentedinotherpublications[15,24].
4 3DOBJECTMODELINGUSINGDATAFROM
THELASERSCANNER
Measurement data of the modeled object were
obtained by the pulse laser scanner Leica
ScanStation. Additional information concerning the
distribution
of the rooms was obtained from the
technical documentation of the building made
availableforscientificandresearchpurposesbythe
Department of Exploitation of the University of
Warmia and Mazury in Olsztyn. Processing of the
results of measurement was performed using the
LeicaCyclone software. Plans and sections of the
building were made inthe LibreCADsoftware. For
3D modeling was used the TrimbleSketchUp
software. Basics of using the program and the
feasibilityofthe3Dmodelperformanceispresented
in detail in the paper [17]. Finally, the constructed
model was compared with the source point cloud
usingtheCloudComparesoftware.
Fieldworkbeganwiththedevelopmentofplanof
the measurement. 4 locations of measuring stations
were determined, ensuring the appropriate
resolutionofapointcloud.Onselectedelementsof
technicalinfrastructure were setrotarytargetplates
enablingsubsequentregistration(joining)processof
point clouds measured from individual stations.
These
plates have very good laser beam reflection
parameters.Atthesametimewasprepared asketch
showing the location and identifiers of individual
plates. The measurement at each station included
scanning the object and target plates.At the end of
field operations was drawn photographic
documentationoftheobjectwiththe
camera.
The next step was to combine all scans into a
single object, namely registration. All point clouds
wereindicatedanda systemofcoordinatesofoneof
them was selected as a global system for which all
theothersweretransformed.Aspointsofadjustment
ofthethree‐dimensional
pointcloudstransformation
they used previously designated centers of target
plates.Unificationofpointcloudswasconducted,in
which process the repetitive points were removed,
filtered to the desired resolution and the numeric
recording of the point cloud was optimized.
Subsequently, the cloud of points was limited by
selecting the fragment
of space that contained the
consideredbuilding(Fig.6),nexttheplanesofwalls
and cross‐sections of the characteristic points were
generatedandexportedtorasterfiles.
Figure6.Pointcloudofthebuildingfacadeinperspective
view(CloudCompare)[source:ownstudy]
Further work wascarried outin LibreCAD (Fig.
7).Distributionofroomsandcorridorswereobtained
from the technical documentation of the building.
The internal accuracy and consistency of all
componentswereensured.
Figure7.Examplesofprojectionsandcross‐sectionsofthe
building(LibreCAD)[source:ownstudy]
The target stage of work was to develop a 3D
modelofthebuildinginTrimbleSketchUp.Elevation
projections and cross‐sections of the building were
imported,whichthenwerepreciselyarrangedinthe
programspace(Fig.8).
Figure8.Plansandcross‐sectionsofthebuilding(Trimble
SketchUp)[source:ownstudy]
Then, based on the geometry of the vector
projectionsandcross‐sectionswaspreparedathree‐
dimensional model of the building (Fig. 9). The
model includes approxima tely 5200 linear and
305
surface elements. The colors of individual pieces
have been provided on the basis of photographic
documentation.
Figure9. 3D model of the building (TrimbleSketchUp)
[source:ownstudy]
Thefigurebelowpresentsselectedelementsofthe
3Dmodelofthemodeledobjectinterior.
Figure10. Selected elements of the building interior
(TrimbleSketchUp)[source:ownstudy]
In order to assess the accuracy of the 3D model
they compared the constructed three‐dimensional
modeloftheobject(externalfacade)fromthesource
point cloud measured by a laser scanner. The
analysiswascarriedoutinCloudCompare(Fig.11).
Theresultsarepresentedintheformofahistogram
ofmatchingerrors (Fig.11,right)thatwereapplied
intheformofcolorsonthemodeloftheobject(Fig.
11,left).
Assuming that the cloud of points is the most
faithful discrete representation of the building
geometry it can be assumed that the error of the
developedfacade
modelisbelow5cm.
Figure11. Evaluation of the accuracyof the model of the
building facade from a point cloud (CloudComapre)
[source:ownstudy]
5 CONCLUSIONS
The process of navigation can be carried out both
outdoorsaswellasinsideobjects.Thedevelopment
of modern mobile devices and their applications
make it possible in the near future to have a very
rapiddevelopmentofpositioningsystems,alsointhe
buildings, and the presentation of objects
of the
navigation infrastructure will be widely
implementedin3D.
The article presents several methods of
implementation of 3D models of objects using
various sources of information and commonly
availablesoftware.Thepaperalsoshowsamethodof
modeling the interior of a typical object with
dimensions of 30.8m ×
11.6m, located at the
UniversityofWarmiaandMazuryinOlsztyn.There
is no doubt that such projects will be in the future
used in navigation systems operating both inside
buildings, as well as they will complement the
database of objects in the navigation systems
workingoutdoors.
REFERENCES
[1]AlkanR.M.,KarsidagG.,Analysisof the Accuracyof
Terrestrial LaserScanningMeasurements,Proceedings
oftheFIGWorkingWeek,Rome,2012.
[2]Arun K. S., Huang T. S., Blostein S. D., Least‐Squares
Fitting of Two 3‐D Point Sets, IEEE Transactions on
Pattern Analysis and Machine Intelligence, Vol.
9(5),
pp.698–700,1987.
[3]BHMW, Znaki, skróty, terminologia stosowane na
mapach wydawanych przez BHMW, Nr 551, Gdynia,
2012(inPolish).
[4]BoehlerW.,VicentM.B.,MarbsA.,InvestigatingLaser
Scanner Accuracy, The International Archives of the
Photogrammetry, Remote Sensing and Spatial
Information Sciences, Vol. 34, Part5/C15, pp.
696‐701,
2003.
[5]Ciećko A., Oszczak S., Grzegorzewski M.,Ćwiklak J.,
Wyznaczenie pozycji statku powietrznego oraz
dokładności z wykorzystaniem systemu EGNOS w
Polsce wschodniej, Logistyka, Nr 6, str. 617‐624, 2010
(inPolish).
[6]Clark J., Robson S., Accuracy of Measurements Made
with a Cyrax 2500 Laser
Scanner Against Surfaces of
Known Colour, Survey Review, Vol. 37(294), pp. 626‐
638,2004.
[7]Czaplewski K., Podstawy nawigacji morskiej i
śródlądowej,WydawnictwoBernardinum,Pelplin,2014
(inPolish).
[8]Ebrahim M. A., 3D Laser Scanners: History,
Applications, and Future, Civil Engineering
Department,FacultyofEngineering,AssiutUniversity,
2011.
[9]
FranceschiM.,TezaG., PretoN., PesciA.,Galgaro A.,
Girardi S., Discrimination Between Marls and
LimestonesUsingIntensityDatafromTerrestrialLaser
Scanner,ISPRSJournalofPhotogrammetryandRemote
Sensing,Vol.64(6),pp.522‐528,2009.
[10]Gajderowicz I., Odwzorowania kartograficzne.
Podstawy, Wydawnictwo Uniwersytetu Warmińsko‐
MazurskiegowOlsztynie,
2009(inPolish).
[11]Hejbudzka K., Lindenbergh R., Soudarissanane S.,
HummeA.,Influenceofatmosphericconditionsonthe
rangedistanceandnumberofreturnedpoints inLeica
ScanStation 2 point clouds, International Archives of
Photogrammetry, Remote Sensing and Spatial
InformationScience,Vol.XXVIII,Part5,CommissionV
306
Symposium, Newcastle upon Tyne, UK, pp. 282‐287,
2010.
[12]KerstenT.P.,SternbergH.,MechelkeK.,Investigations
into the Accuracy Behaviour of the Terrestrial Laser
Scanning System Trimble GS100, Optical 3D
MeasurementTechniquesVII,Vol.I,pp.122‐131,2005.
[13]KopaczP.,WeintritA.,GeodesicBasedTrajectoriesin
Navigation
Coordinates, The Monthly Magazine on
Positioning,NavigationandBeyond,Vol.VII(6),pp.32‐
35,2011.
[14]Lindstaedt M., Graeger T., Mechelke K., Kersten T.,
TerrestrischeLaserscannerimPrüfstand–Geometrische
Genauigkeitsuntersuchungenaktuellerterrestrischer
Laserscanner, Photogrammetrie, Laserscanning,
Optische3D‐Messtechnik,pp.4‐14,2011(inGerman).
[15]Lindstaedt M., Kersten T., Mechelke K., Graeger
T.,
Prüfverfahrenfürterrestrische Laserscanner –
Gemeinsamegeometrische
Genauigkeitsuntersuchungenverschiedener
Laserscanner an der HCU, Photogrammetrie,
Laserscanning, Optische 3D‐Messtechnik, pp. 264‐275,
2012(inGerman).
[16]MacEachren A. M., Johnson G. B., The Evolution,
Application and Implications of Strip Format Travel
Maps, The Cartographic Journal, Vol. 24, pp. 147‐158,
1987.
[17]OszczakB.,Tanajewski
D.,HarmacińskiA.,Klimczuk
M., Modelowanie trójwymiarowe budynków lotniska
Dajtki‐Olsztyn w aplikacjach AutoCAD Civil 3D i
GoogleSketchUp,RocznikiGeomatyki,TomIX,Zeszyt
4(48),str.129‐137,2011(inPolish).
[18]Resolution of Minister of Administration and
Digitalization of 2 November 2015 on Database of
Topographic Objects and
Base Map, Dz.U. 2015 poz.
2028,2015.
[19]Rudolf S., 10 Years of Terrestrial Laser Scanning–
Technology, Systems and Applications, 2011,
https://www.fig.net/news/news_2011/geosiberia_april_
2011/Novosibirsk_Geosiberia_2011_Paper_Rudolf_Staig
er.pdf.Datadostępu(Accessed5.12.2015).
[20]Slob S., Hack R., 3D Terrestrial Laser Scanning as a
New Field Measurement and Monitoring Technique,
Engineering Geology for Infrastructure Planning in
Europe,LectureNotesinEarthSciences,Springer,Vol.
104,pp.179‐189,2004.
[21]Specht C., Dąbrowski P., Specht M., Koc W.,
Chrostowski P., Szmagliński J., Dera M., Skóra M.,
Mobilne pomiary satelitarne na liniach Pomorskiej
KoleiMetropolitalnej,PrzeglądKomunikacyjny,Tom7,
str.9‐16,2016(in
Polish).
[22]Specht C., Deniziuk M., Kamiński M., Wykorzystanie
oprogramowania SketchUp jako narzędzia do
tworzenia modeli 3D obiektów na potrzeby
symulatorów mostka nawigacyjnego, Prace Wydziału
NawigacyjnegoAkademiiMorskiejwGdyni,Nr29,str.
46‐54,2014(inPolish).
[23]Specht C., Szot T., Specht M., Badanie dokładno
ści
personalnych odbiorników GPS w pomiarach
dynamicznych, TTS Technika Transportu Szynowego,
Nr10,str.2547‐2555,2013(inPolish).
[24]Sternberg H., Kersten T., Comparison of Terrestrial
Laser Scanning Systems in Industrial As‐Built‐
DocumentationApplications,Optical3DMeasurement
TechniquesVIII,Vol.I,pp.389‐397,2007.
[25]U.S. DoD,
Global Positioning System Standard
PositioningServicePerformanceStandard,3rdEdition,
2001.
[26]Vozikis G., Haring A., Vozikis E., Kraus K., Laser
Scanning: A New Method for Recording and
DocumentationinArchaeology,ProceedingsoftheFIG
WorkingWeek,Athens,2004.
[27]Yilmaz H. M., Yakar M., Yildiz F., Digital
Photogrammetry in Obtaining
of 3D Model Data of
Irregular Small Objects, The International Archives of
the Photogrammetry, Remote Sensing and Spatial
Information Sciences, Vol. 37, Part B3b, pp. 125‐130,
2008.
[28]Weintrit A., Development of the IMO e‐Navigation
Concept –CommonMaritime DataStructure. Modern
Transport Telematics, Communications in Computer
and
Information Science, Springer, Vol. 239, pp. 151‐
163,2011.
[29]WeintritA.,Morskiepomocenawigacyjne –definicja,
podział i klasyfikacja, Przegląd Telekomunikacyjny +
WiadomościTelekomunikacyjne,Nr 11,str.1656‐1660,
2012(inPolish).
