Ci a ion: Aska , C.; S e nbe g, H. Use
o Sma phone Lida Technology o
Low-Cos 3D Building
Documen a ion wi h iPhone 13 P o:
A Compa a i e Analysis o Mobile
Scanning Applica ions. Geoma ics
2023,3, 563–579. h ps://doi.o g/
10.3390/geoma ics3040030
Academic Edi o : Pie e
G ussenmeye
Recei ed: 12 Augus 2023
Re ised: 8 Decembe 2023
Accep ed: 8 Decembe 2023
Published: 11 Decembe 2023
Copy igh : © 2023 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
A icle
Use o Sma phone Lida Technology o Low-Cos 3D Building
Documen a ion wi h iPhone 13 P o: A Compa a i e Analysis o
Mobile Scanning Applica ions
Cigdem Aska * and Ha ald S e nbe g
Depa men o Geoma ics, Ha enCi y Uni e si y, 20457 Hambu g, Ge many; ha ald.s e nbe g@hcu-hambu g.de
*Co espondence: cigdem.aska @g .hambu g.de
Abs ac :
Lase scanning echnology has long been he p e e ed me hod o cap u ing in e io scenes
in a ious indus ies. Wi h a g owing ma ke , smalle and mo e a o dable scanne s ha e eme ged,
o e ing end p oduc s wi h su icien accu acy. While no on pa wi h p o essional scanne s, Apple
has made lase scanning echnology accessible o use s wi h he in oduc ion o he new iPhone P o
models, democ a izing 3D scanning. Thus, his s udy aimed o assess he pe o mance o he iPhone’s
lida echnology as a low-cos solu ion o building documen a ion. Fou scanning applica ions we e
e alua ed o de e mine he accu acy, p ecision, and use expe ience o he gene a ed poin clouds
compa ed wi h a e es ial lase scanne . The esul s e eal a ying pe o mances on he same
de ice, highligh ing he in luence o so wa e. No ably, he e is oom o imp o emen , pa icula ly in
acking he de ice’s posi ion h ough so wa e solu ions. As i s ands, he echnology is well sui ed
o applica ions such as indoo na iga ion and he gene a ion o quick loo plans in he con ex o
building documen a ion.
Keywo ds:
low-cos solu ions; sma phone lida ; iPhone 13 P o; building documen a ion; poin
cloud analysis
1. In oduc ion
O e he pas decades, lase scanning has eme ged as a cu ing-edge echnology.
Lase scanne s gene a e poin clouds ha a e highly e ec i e in ep esen ing objec s o
a ying complexi y a di e en scales [
1
]. In he 1990s, e es ial lase scanne s (TLS)
we e in oduced o he su eying indus y [
2
], and owa ds he 2010s, hey became mo e
accu a e and capable o scanning anges o hund eds o me e s. TLSs a e widely used in
a a ie y o applica ions, including cul u al he i age [
3
,
4
], change de ec ion [
5
,
6
], moni o ing
and de o ma ion [
7
–
9
], as-buil modelling [
10
], and o es y [
11
]. In he la e 2000s, mobile
mapping sys ems (MMS), which ope a e on a ehicle such as a ca , we e in oduced
in o mapping ope a ions, mainly o da a cap u e on oad in as uc u e and building
acades [12] and ex ended i s use o a ious applica ions [13–20].
These sys ems u ilize ac i e o passi e sensing o cap u e he objec o in e es , along
wi h GNSS and IMU o accu a e geo e e encing. While he GNSS and IMU combina ion
wo ks well o ou doo applica ions, in GNSS-denied spaces like indoo s, using only ine ial
senso s leads o an inc easing d i a e, one which canno be co ec ed due o he unknown
unc ion wi h espec o ime [
21
]. Simul aneous localiza ion and mapping (SLAM) is one o
he echniques ha o e s a solu ion o his p oblem. I s undamen al concep is moni o ing
he senso ’s posi ion and o ien a ion (pose) o e ime in 3 deg ees o eedom (DoF) and
wi h ela i e coo dina es, espec i ely. This is achie ed by u ilizing o e laps in op ical da a,
such as wi h p e iously obse ed ea u es [21].
Nowadays, nume ous low-cos MMS ely on SLAM and can be u ilized h ough
a ious pla o ms like olleys, backpacks, and hand-held de ices. Al hough many o
hese sys ems ha e been ini ialized o en e ainmen , some ha e led o esea ch wo k
Geoma ics 2023,3, 563–579. h ps://doi.o g/10.3390/geoma ics3040030 h ps://www.mdpi.com/jou nal/geoma ics
Geoma ics 2023,3564
de elopmen s o u he applica ions. In addi ion o mobile lase scanne solu ions, dep h
came as ep esen ano he commonly employed low-cos al e na i e in 3D documen a ion.
The in eg a ion o RGB and dep h came as gene a es a 3D ep esen a ion o he scene
by cap u ing he dis ance be ween he objec and he came a wi hin hei ield o iew
(FOV) and is equen ly u ilized in compu e ision [
22
]. Two common app oaches o
dep h came as a e ime-o - ligh (ToF) and s uc u ed ligh . ToF came as, exempli ied by
de ices like Azu e Kinec and HoloLens, emi ligh pulses and cap u e he e lec ed signal o
calcula e he dis ance based on he measu ed ime o he ligh o a el o an objec and back.
Nume ous s udies ha e inco po a ed bo h sys ems in indoo mapping [
23
–
25
]. S uc u ed
ligh -based came as p ojec a known ligh pa e n on o he scene and calcula e dep h
in o ma ion based on he dis o ion o he pa e n on he analyzed objec su ace. Ea ly
gene a ions o Kinec se e as a well-known example o his ype o came a and ha e been
u ilized in a ious s udies o in es iga e hei capabili ies in indoo mapping [22,26,27].
The de elopmen s in lase scanning echnology and he apid ad ancemen in low-
cos senso echnology ha e made 3D lase scanning mo e accessible and cos -e ec i e.
O e he yea s, esea che s ha e in es iga ed compa a i e e alua ion o he lida -based
indoo MSS, such as [
28
–
31
]. E en consume echnology, like some iPhone models, now
inco po a es lase scanning echnology, opening possibili ies o he democ a iza ion o
3D scanning. This pape aims o pe o m scanning expe imen s wi h Apple iPhone 13
P o lida o 3D documen a ion o indoo en i onmen s. People spend mos o hei ime
in indoo en i onmen s, [
32
] ye hese lack p ope and up- o-da e map ep esen a ions.
Though de elopmen s in scanning echnology ha e made i possible o cap u e indoo
en i onmen s wi h e iciency o ime and accu acy, he cos could s ill be lowe and he
echnology needs expe ise. The e o e, esea ch in o low-cos oppo uni ies o indoo
mapping, as in o he domains [33], is s ill an ongoing e o .
In his ega d, his pape will assess he possibili y o using a consume -g ade sma -
phone equipped wi h lida (Apple iPhone 13 P o) as a low-cos al e na i e o mobile
mapping sys ems o e es ial lase scanne s in he 3D documen a ion o indoo en i on-
men s, such as in he quick gene a ion o loo plans and indoo na iga ion maps, de ec ing
changes in spaces, o illing he gaps in a p e ious scan. Fo he expe imen , a oom occu-
pied by labo a o y in en o y will be scanned by di e en 3D scanning applica ions ins alled
on an iPhone 13 P o, and he esul ing poin clouds will be compa ed wi h e es ial lase
scanne da a.
The emainde o his pape is o ganized as ollows. Sec ion 2b ie ly desc ibes he
key ela ed wo ks on sma phone-based MSS and exis ing solu ions. Sec ion 3explains he
me hodologies used as well as he da a acquisi ion. The esul s a e exp essed in Sec ion 4,
and, inally, he pape is concluded and discussed in Sec ion 5.
2. Rela ed Wo ks
Using sma phones o ob ain spa ial in o ma ion is no a new concep , as sma phones
a e equipped wi h ine ial senso s ha a e commonly used in indoo posi ioning, such as
in [
34
], and came as ha a e used in 3D econs uc ion based on images o ideos [
35
,
36
].
Mos ea lie s udies in ensi ely wo ked wi h he Google Tango echnology, which was
launched in 2014 [
37
] and aimed o e alua e he dependabili y, in luence, and engage-
men o use s in a ha dwa e and so wa e bundle ha pe mi s he de elopmen o aug-
men ed/mixed/ i ual eali y con en exclusi ely h ough he use o hei sma phones
o able s [
38
]. The Tango p ojec was only a ailable on a limi ed numbe o compa ible
phones and able s. In 2018, he p ojec was e mina ed and eplaced wi h ARCo e [
39
].
Some s udies include [
40
,
41
], bo h o which es ed he Tango able ’s capabili y o 3D docu-
men a ion o indoo spaces. O he examples a e [
42
], which in es iga ed 3D econs uc ion
using a Tango sma phone in he con ex o cul u al he i age, and [
38
], which assessed he
quali y and po en ial o he sys em in hei s udy.
Apple in oduced lida senso s in o i s p o lines o able s and sma phones, iPad P o
and iPhone 12 P o, in 2020. This b ough a no el y o he 3D scanning subjec by inco po-
Geoma ics 2023,3565
a ing a lida senso in o use -g ade sma phones, leading o he ques ion o whe he hese
de ices would be a low-cos al e na i e wi h enough accu acy in 3D scanning. Apple’s
aim was mo e o imp o e he came a and enhance he augmen ed eali y expe ience o
i s use s. Hence, Apple has no eleased any 3D scanning applica ions o la ge spaces
o objec s a e he ini ial elease, apa om he Measu e app, which is designed as
a measu ing ool. Howe e , Apple has p o ided a so wa e de elopmen ki (SDK); since
hen, many de elope s ha e de eloped 3D scanning apps wi h ARKi by Apple. As i
seems o be compa ible wi h no ice use s who seek o gene a e a loo plan o design hei
houses o o y u ni u e be o e buying, mo e applica ions ha a ge scanning expe s
ha e been eleased o e ime. I has also ecei ed a en ion om esea che s as a low-cos
and o e - he-shel al e na i e o 3D documen a ion. Di e en subjec s ha e been in es-
iga ed since he elease o he i s Apple de ice equipped wi h he lida senso . [
43
,
44
]
e alua ed he iPhone 12 P o o i s use in geoscience applica ions. The o me epo s a 10
cm senso accu acy when demons a ing i s use on a coas al cli , while he la e concludes
ha he es ed iPhone 12 P o de ice would be he s anda d p ocess o cap u ing ocky
slopes and in es iga ing discon inui ies, despi e limi a ions in i s ange. [
33
,
45
] assessed
he Apple lida de ices o hei use in he i age documen a ion and concluded ha his
echnology holds g ea p omise o he nea u u e. [
46
] in es iga ed hese de ices o
indoo /ou doo modelling and epo ed 53 cm o local p ecision and 10 cm o global
co ec ness. The indoo es space consis ed o wo adjacen ooms ha co e ed a o al o
a ound 200 m
2
. [
47
] e alua ed he iPad P o om he a chi ec u al su eying pe spec i e
and epo ed 2 cm p ecision and 4 cm accu acy o a 1:200 map scale.
3. Ma e ials and Me hods
The iPhone 13 P o was he de ice es ed in his s udy, and was eleased in Sep embe
2021. The de ice weighs 204 g, has a 7.7 mm hickness, and ea u es a 6.1-inch supe e ina
display. I is powe ed by an A15 Bionic chip wi h a 6 co e CPU, 5-co e GPU, 16-co e Neu al
Engine, 6 GB RAM, and 128 GB memo y. Addi ionally, he iPhone 13 P o includes h ee
12MP ea came as ( elepho o, wide, and ul awide) and a 3D ime-o - ligh (ToF) lida .
Al hough Apple publishes limi ed in o ma ion abou he echnical de ails o he lase used
in hei p oduc s, he au ho s o [
34
] ha e claimed ha he lase senso is a solid-s a e
de ice ha does no use mo o ized mechanical pa s so as o p o ide highe scalabili y and
eliabili y. Acco ding o [
27
], he lida senso o he iPhone 13 P o emi s a e ical ca i y
su ace emi ing lase wi h di ac ion op ics elemen (VCSEL DOE) a a nea -in a ed
spec um in a 2D a ay and is ecei ed by a single-pho on a alanche pho odiode (SPAD).
A o al o 576 poin s a e emi ed in an a ay o 8 ×8 poin s, di ac ed in o 3 ×3 g ids.
Al hough Apple does no o e a dedica ed 3D scanning applica ion, de elope s
can access senso s on iOS 14 and la e e sions h ough ARKi o c ea e 3D mapping
applica ions. As a esul , se e al 3D scanning applica ions a e a ailable in he Apple S o e.
This s udy used ou di e en 3D scanning applica ions—3DScanne , PolyCam, Scani e se,
and Si eScape. The selec ion was based on h ee c i e ia: (1) he applica ion was ee o
had a ee-use op ion, (2) he p oduc gene a ed a poin cloud, and (3) he lida senso
was u ilized in poin cloud gene a ion. Each applica ion is explained in he ollowing
subsec ions, and a summa y o he applica ions’ speci ica ions is gi en in Table 1below.
3.1. 3D Scanne App
The 3D Scanne app ( e sion 2.0.13(1)) is a ee applica ion o e ing mul iple scan
modes, including LIDAR, LIDAR Ad ance, Poin Cloud, RoomPlan, Pho os, and T ueDep h.
The applica ion’s help page explains each mode o help use s selec he bes mode o he
scanne ’s pu pose. Al hough he LIDAR Ad ance mode o e s lexibili y in se ing pa-
ame e s ( esolu ion, max dep h, e c.) be o e he scan, he LIDAR mode was used in his
s udy as sugges ed o la ge a eas. In he ad anced mode, he scan au oma ically ends a e
a sho cap u e ime due o he la ge numbe o poin s, while he LIDAR mode enables
longe scans. The quali y p oduced by bo h modes is epo ed o be he same. Once he
Geoma ics 2023,3566
cap u e is comple ed, he scan is p ocessed (smoo hing, simpli ying, and ex u ing) in HD,
as , o cus om modes. The app includes ex a ea u es such as ex ending a scan, iewing
he came a ajec o y, measu ing wi h he scan, and cap u ing a loo plan image. Expo s
a e ei he poin cloud (PCD, PLY, LAS, e57, PTS, XYZ) o mesh (OBJ, KMZ, FBX, e c.). The
LAS o ma expo s geo e e enced poin clouds wi h he WGS84 coo dina es. The scanned
da a we e expo ed in XYZ o ma , compa ible wi h he poin cloud p ocessing so wa e
CloudCompa e ( e sion. 2.12.4).
Table 1.
Summa y o he speci ica ions o each applica ion. Gi en in o ma ion is based on he used
e sions a he ime o da a cap u e. By he ime o he publica ion o he pape , he e migh be
changes in he speci ica ions.
3D Scanne App PolyCam Si eScape Scani e se
Scan mode
LIDAR, LIDAR
Ad ance, Poin Cloud,
Pho os, T ueDep h
LIDAR, Pho o, Room LIDAR Small objec , medium
objec , la ge objec (a ea)
Scan se ings Resolu ion, max dep h - Poin densi y and size
(low, med, high) Range se ing (max 5 m)
P ocessing
op ions HD, Fas , Cus om Fas , Space, Objec ,
Cus om
Synching o he
Si eScape cloud Speed, a ea, de ail
P ocessing s eps Smoo hing,
simpli ying, ex u ing - - -
Expo as Poin cloud, mesh Poin cloud, mesh Poin cloud Poin cloud, mesh
Expo o ma s
PCD, PLY, LAS, e57,
PTS, XYZ, OBJ, KMZ,
FBX e c.
DXF, PLY, LAS, PTS,
XYZ, OBJ, STL, FBX e c.
e57 PLY, LAS, OBJ, FBX, STL,
GLB, USDZ
3.2. PolyCam
PolyCam ( e sion 3.0.2) o e s ee, eam (14.99 $/sea ), and p o (14.99 $/mon h)
e sions. The ee e sion was su icien o his s udy as i does no limi lida cap u es;
howe e , he ee ial e sion was used o he ease o da a expo . The scan modes a ail-
able a e LIDAR, pho o, and oom. The pho o mode uses he pho og amme y echnique
and is sui able o smalle objec s, while he oom mode gene a es 3D models ins an ly.
Cap u ed scans can be p ocessed unde as , space, objec , o cus om ca ego ies. Mea-
su emen s on scans and ex ending o edi ing an exis ing scan a e possible. Scans can be
expo ed as poin clouds (DXF, PLY, XYZ, LAS, PTS) o mesh (OBJ, FBX, STL, e c.). This
wo k used LIDAR mode o da a cap u e using he PolyCam applica ion, and he ou pu
was expo ed in XYZ o ma .
3.3. Si eScape
Si eScape ( e sion 1.6.9) also o e s ee, eam (N/A p ice) and p o (49.99$/mon h–
52.99
€
) e sions. Up o 50 sqm is included in he ee e sion, wi h one scan synced o hei
web iewe . Expo is limi ed o PLY o E57 o ma s in he ee e sion. Si eScape wo ks
only in LIDAR mode. The use can se poin densi y (low, med, o high), which a ec s how
quickly he scan will each he maximum allowed poin , and poin size (low, med, o high),
which only se s he displaying size o he poin s while scanning. A e app oxima ely one
minu e, he maximum poin limi was eached o one scan, and up o en scans could be
cap u ed consecu i ely. The comple ed scans can be expo ed as a poin cloud o synched
o he Si eScape cloud o iewing in a web app o sha ing wi h mul iple use s. The Geolab
cap u e was comple ed wi h he medium (med) poin densi y se ing o Si eScape and as
en pa ial scans. These scans we e conduc ed consecu i ely, u ilizing some o e lapping
a eas in be ween. The scans we e expo ed in E57 o ma .
Geoma ics 2023,3567
3.4. Scani e se
Scani e se ( e sion 2.0.3) is a ee applica ion ha o e s scan modes based on he size
o he objec (small, medium, la ge). The p ocessing is a ailable in speed, a ea, and de ail
modes. P ocessed scans can be expo ed as a poin cloud (PLY, LAS) o mesh (OBJ; FBX,
STL, GLB, USDZ). The LAS o ma allows he expo ing o poin clouds geo e e enced wi h
UTM Ca esian coo dina es, which we e used in his wo k. The scans we e comple ed in
la ge objec mode and p ocessed in a ea mode.
Da a we e collec ed in he geoma ics labo a o y (Geolab) a Ha enCi y Uni e si y,
Hambu g (Figu e 1). The Geolab is an ideal loca ion o es ing he capaci y o he iPhone’s
lida senso in a con olled space. I has a 35 m long s aigh conc e e wall on one o i s
longe edges. The o he long edge comp ises wo walls measu ing 13 and 23 m in leng h,
which g adually widen owa ds he cen e and connec wi h each o he . These walls ha e
la ge windows co e ing hem. The b oade windows we e cu ained be o e scanning. The
sho side walls a e 7 and 9 m long. The e a e six su eying pilla s ha a e app oxima ely
1.5 m high and 40 cm in size, as well as many lase scanning a ge s, some o which had
been p e iously measu ed wi h a o al s a ion. Addi ionally, he Geolab is clu e ed wi h
u ni u e and equipmen .
Geoma ics 2023, 3, FOR PEER REVIEW 6
Figu e 1. The images on he le (a–c) illus a e he Geolab es oom. The whole oom was scanned
a Pa 1 and Pa 2, co e ing a common a ea as shown on he igh side (d). The scan concludes a
he s a ing poin , iden i ied as a poin on he image (d). The colo ed a ows (d) indica e he walking
di ec ion du ing he scanning p ocess.
4. Resul s
Upon da a collec ion, all da a p ocessing o each applica ion was conduc ed using
he open-sou ce poin cloud p ocessing so wa e CloudCompa e [49]. Fi s , he poin
clouds o Pa 1 and Pa 2 o each applica ion we e oughly aligned wi h he TLS cloud
by u ilizing exis ing lase scanning a ge s o o he dis inc i e poin s. Nex , he i e a i e
closes poin (ICP) algo i hm pe o med a ine egis a ion on each pa . Regis e ed pa s
we e hen me ged o gene a e a single poin cloud o he es oom o each applica ion.
Figu e 2 displays he egis e ed poin clouds o each applica ion. PolyCam appea s o
ha e less dis o ion compa ed wi h he o he applica ions, which, o example, exhibi
mo e dis o ed edges. The 3D Scanne app has some a eas on he ceiling ha we e no
cap u ed, which was a esul o missing cap u e. The Si eScape poin cloud has a e y high
numbe o poin s, 115,883,552, in compa ison wi h PolyCam (6,685,940), 3D Scanne app
(6,568,595), and Scani e se (787,819). On one o he la wall su aces, he poin densi y
was assessed wi hin a one m2 box. The poin dis ibu ion was as ollows: 204,231 poin s
o Si eScape, 9128 poin s o PolyCam, 6,405 poin s o 3D Scanne app, and 1183 poin s
o Scani e se.
Figu e 1.
The images on he le (
a
–
c
) illus a e he Geolab es oom. The whole oom was scanned
a Pa 1 and Pa 2, co e ing a common a ea as shown on he igh side (
d
). The scan concludes a
he s a ing poin , iden i ied as a poin on he image (
d
). The colo ed a ows (
d
) indica e he walking
di ec ion du ing he scanning p ocess.

Geoma ics 2023,3568
Each applica ion is used o scan he Geolab in wo pa s (Figu e 1) by c ea ing a loop
o each pa (excep Si eScape). Scanning is epea ed a numbe o imes, and he op imal
esul was achie ed when he phone was held pa allel o he walls and mo ed up and
down by sliding slowly owa d one side a e e y s ep. A en ion is gi en o main aining
he dis ance be ween he scanned su ace and he came a, ensu ing ha i does no exceed
5 m, as ecommended by he applica ions. E o s we e made o co e he ceiling and
he loo en i ely while adhe ing o he ecommenda ions in he applica ions’ manuals by
a oiding apid mo emen s and sudden u ns. The scanning ime o each applica ion was
simila , aking be ween 20 o 25 min o cap u e he en i e oom. Fu he mo e, he labo a o y
was scanned wi h he e es ial lase scanne (TLS) Z+F Image 5016 [48] om eigh scan
posi ions, and The TLS da a se ed as e e ences in e alua ion.
This pape in es iga es he capaci ies o he iPhone 13 P o lida as a low-cos sen-
so al e na i e o 3D documen a ion o indoo en i onmen s, wi h a ocus on he qual-
i y o he senso and he gene a ed poin cloud. The global accu acy o he gene a ed
poin clouds was e alua ed by compa ing hem o he e es ial lase scanne da a using
a cloud- o-cloud me hod, and segmen ed planes we e analyzed o assess he p ecision o
he senso . Dis ances we e calcula ed o de e mine he local accu acy o he sys em by using
al eady a ailable a ge s in he Geolab. The use o di e en applica ions in he e alua ion
aims o e eal he e ec o he so wa e on quali y o he inal poin cloud. Finally, he use
expe ience is included in he discussion and conclusion sec ions o he e alua ion.
4. Resul s
Upon da a collec ion, all da a p ocessing o each applica ion was conduc ed using he
open-sou ce poin cloud p ocessing so wa e CloudCompa e [
49
]. Fi s , he poin clouds o
Pa 1 and Pa 2 o each applica ion we e oughly aligned wi h he TLS cloud by u ilizing
exis ing lase scanning a ge s o o he dis inc i e poin s. Nex , he i e a i e closes poin
(ICP) algo i hm pe o med a ine egis a ion on each pa . Regis e ed pa s we e hen
me ged o gene a e a single poin cloud o he es oom o each applica ion. Figu e 2
displays he egis e ed poin clouds o each applica ion. PolyCam appea s o ha e less
dis o ion compa ed wi h he o he applica ions, which, o example, exhibi mo e dis o ed
edges. The 3D Scanne app has some a eas on he ceiling ha we e no cap u ed, which
was a esul o missing cap u e. The Si eScape poin cloud has a e y high numbe o
poin s, 115,883,552, in compa ison wi h PolyCam (6,685,940), 3D Scanne app (6,568,595),
and Scani e se (787,819). On one o he la wall su aces, he poin densi y was assessed
wi hin a one m
2
box. The poin dis ibu ion was as ollows: 204,231 poin s o Si eScape,
9128 poin s o PolyCam, 6,405 poin s o 3D Scanne app, and 1183 poin s o Scani e se.
A close look a he poin clouds shows some spli su aces, pa icula ly whe e wo
pa s o e lap and loops end. These s em om he d i e o accumula ing o e ime and
a e a known p oblem in SLAM sys ems. Figu e 3shows examples o he spli su aces in
each applica ion’s poin cloud.
The compa ison ini ially assesses global accu acy using he mul iscale model- o-model
cloud compa ison (M3C2) me hod [
5
]. This me hod calcula es he Euclidean dis ance
be ween poin clouds along he su ace, ypically o a speci ied sea ch dep h. The esul ing
de ia ions om he e e ence poin cloud a e ep esen ed as M3C2 dis ances on he colo -
coded cloud. Fo his s udy, a sea ch dep h o 40 cm was used, conside ing dis ances
beyond his alue as useless.
Figu e 4p esen s he esul s, whe e a ange o 40 cm and a colo sa u a ion o 14 cm
we e used. The 3D Scanne app shows highe de ia ions on he walls compa ed wi h he
loo and ceiling. PolyCam demons a es o e all balanced and low de ia ions, wi h some
peaks obse ed on he loo and ceiling. Simila ly, Si eScape displays balanced de ia ions,
bu he walls expe ience pa ially highe de ia ions. Scani e se does no display ema kable
pe o mance in any speci ic a ea, bu he walls show smalle de ia ions compa ed wi h
he loo and ceiling. Despi e employing a consis en scanning app oach wi h he phone
held pa allel o he side walls and mo ed up and down by he same use a a no mal o
Geoma ics 2023,3569
slow pace, a ying pe o mance in di e en a eas wi hin he es oom is a ibu ed o
he SLAM algo i hm. A disce nible line e eals he ope a o ’s pa h as hey walked along
one wall in one di ec ion and hen back along he o he wall and can be obse ed as a
sligh o dominan line on he poin clouds (Figu e 4). Addi ionally, a eas whe e he loops
end, o he di e en pa s o he oom a e connec ed exhibi highe de ia ions ac oss all
applica ions. These obse a ions highligh he consequences o e o s in pose es ima ion
du ing dynamic scanning, esul ing in misaligned poin s and ailu e in loop closu e.
Geoma ics 2023, 3, FOR PEER REVIEW 7
Figu e 2. Regis e ed and me ged poin clouds om each applica ion.
A close look a he poin clouds shows some spli su aces, pa icula ly whe e wo
pa s o e lap and loops end. These s em om he d i e o accumula ing o e ime and
a e a known p oblem in SLAM sys ems. Figu e 3 shows examples o he spli su aces in
each applica ion’s poin cloud.
Figu e 3. Example o spli su aces om da ase . Red boxes illus a e spli walls and une en su aces
on he loo o ceiling.
The compa ison ini ially assesses global accu acy using he mul iscale model- o-
model cloud compa ison (M3C2) me hod [5]. This me hod calcula es he Euclidean dis-
ance be ween poin clouds along he su ace, ypically o a speci ied sea ch dep h. The
esul ing de ia ions om he e e ence poin cloud a e ep esen ed as M3C2 dis ances on
he colo -coded cloud. Fo his s udy, a sea ch dep h o 40 cm was used, conside ing dis-
ances beyond his alue as useless.
Figu e 4 p esen s he esul s, whe e a ange o 40 cm and a colo sa u a ion o 14 cm
we e used. The 3D Scanne app shows highe de ia ions on he walls compa ed wi h he
loo and ceiling. PolyCam demons a es o e all balanced and low de ia ions, wi h some
Figu e 2. Regis e ed and me ged poin clouds om each applica ion.
Geoma ics 2023, 3, FOR PEER REVIEW 7
Figu e 2. Regis e ed and me ged poin clouds om each applica ion.
A close look a he poin clouds shows some spli su aces, pa icula ly whe e wo
pa s o e lap and loops end. These s em om he d i e o accumula ing o e ime and
a e a known p oblem in SLAM sys ems. Figu e 3 shows examples o he spli su aces in
each applica ion’s poin cloud.
Figu e 3. Example o spli su aces om da ase . Red boxes illus a e spli walls and une en su aces
on he loo o ceiling.
The compa ison ini ially assesses global accu acy using he mul iscale model- o-
model cloud compa ison (M3C2) me hod [5]. This me hod calcula es he Euclidean dis-
ance be ween poin clouds along he su ace, ypically o a speci ied sea ch dep h. The
esul ing de ia ions om he e e ence poin cloud a e ep esen ed as M3C2 dis ances on
he colo -coded cloud. Fo his s udy, a sea ch dep h o 40 cm was used, conside ing dis-
ances beyond his alue as useless.
Figu e 4 p esen s he esul s, whe e a ange o 40 cm and a colo sa u a ion o 14 cm
we e used. The 3D Scanne app shows highe de ia ions on he walls compa ed wi h he
loo and ceiling. PolyCam demons a es o e all balanced and low de ia ions, wi h some
Figu e 3.
Example o spli su aces om da ase . Red boxes illus a e spli walls and une en su aces
on he loo o ceiling.
Geoma ics 2023,3570
Geoma ics 2023, 3, FOR PEER REVIEW 8
peaks obse ed on he loo and ceiling. Simila ly, Si eScape displays balanced de ia ions,
bu he walls expe ience pa ially highe de ia ions. Scani e se does no display ema k-
able pe o mance in any speci ic a ea, bu he walls show smalle de ia ions compa ed
wi h he loo and ceiling. Despi e employing a consis en scanning app oach wi h he
phone held pa allel o he side walls and mo ed up and down by he same use a a no -
mal o slow pace, a ying pe o mance in diffe en a eas wi hin he es oom is a ibu ed
o he SLAM algo i hm. A disce nible line e eals he ope a o ’s pa h as hey walked along
one wall in one di ec ion and hen back along he o he wall and can be obse ed as a
sligh o dominan line on he poin clouds (Figu e 4). Addi ionally, a eas whe e he loops
end, o he diffe en pa s o he oom a e connec ed exhibi highe de ia ions ac oss all
applica ions. These obse a ions highligh he consequences o e o s in pose es ima ion
du ing dynamic scanning, esul ing in misaligned poin s and ailu e in loop closu e.
Figu e 4. Cloud- o-cloud compa ison o each poin cloud wi h he e e ence TLS poin cloud (de-
pic ed on he le ). Dis ances we e compa ed wi hin a 40 cm ange, wi h de ia ions beyond his
ange esul ing in emp y spaces as obse ed wi hin he ma ked ci cle in Scani e se’s poin cloud.
Highe de ia ions a e seen in diffe en pa s o diffe en poin clouds.
Addi ionally, one o he long side walls was pa ially co e ed by la ge windows,
which we e mos ly shielded om di ec sunligh du ing he cap u e p ocess. The appli-
ca ions, pa icula ly PolyCam and Si eScape, demons a ed sa is ac o y pe o mance
along his wall, sugges ing ha changes in ligh ing condi ions du ing scans had minimal
effec . Howe e , his is an assump ion and no assessed wi hin his wo k. On he o he
hand, he back wall exhibi ed highe de ia ions om he TLS in each applica ion’s poin
cloud. This can be pa ly a ibu ed o he p esence o clu e in on o he wall, which
hinde ed scanning a a close ange. In pa icula , Scani e se expe ienced difficul ies in
his a ea, as indica ed by he dashed ci cle in Figu e 4, whe e i ailed o cap u e any da a
wi hin a 40 cm dis ance om he e e ence cloud.
Table 2 summa izes he esul s ob ained om he isualiza ion in Figu e 4. Si eScape
exhibi s he lowes s anda d de ia ion o 6 cm, ollowed by PolyCam wi h 7 cm,
Figu e 4.
Cloud- o-cloud compa ison o each poin cloud wi h he e e ence TLS poin cloud (depic ed
on he le ). Dis ances we e compa ed wi hin a 40 cm ange, wi h de ia ions beyond his ange
esul ing in emp y spaces as obse ed wi hin he ma ked ci cle in Scani e se’s poin cloud. Highe
de ia ions a e seen in di e en pa s o di e en poin clouds.
Addi ionally, one o he long side walls was pa ially co e ed by la ge windows, which
we e mos ly shielded om di ec sunligh du ing he cap u e p ocess. The applica ions,
pa icula ly PolyCam and Si eScape, demons a ed sa is ac o y pe o mance along his wall,
sugges ing ha changes in ligh ing condi ions du ing scans had minimal e ec . Howe e ,
his is an assump ion and no assessed wi hin his wo k. On he o he hand, he back wall
exhibi ed highe de ia ions om he TLS in each applica ion’s poin cloud. This can be
pa ly a ibu ed o he p esence o clu e in on o he wall, which hinde ed scanning a
a close ange. In pa icula , Scani e se expe ienced di icul ies in his a ea, as indica ed by
he dashed ci cle in Figu e 4, whe e i ailed o cap u e any da a wi hin a 40 cm dis ance
om he e e ence cloud.
Table 2summa izes he esul s ob ained om he isualiza ion in Figu e 4. Si eScape
exhibi s he lowes s anda d de ia ion o 6 cm, ollowed by PolyCam wi h 7 cm, Scani e se
wi h 8 cm, and 3D Scanne app wi h 9 cm. Each applica ion exhibi s i s highes poin
densi y wi hin he 1–3 cm ange, wi h Si eScape leading a 46%, ollowed by PolyCam a
32%, Scani e se a 31%, and he 3D Scanne app a 30%. While Scani e se only has 3% o
poin s alling wi hin he de ia ion ange o 20 o 40 cm, as shown in Figu e 4, i is impo an
o no e ha mos o he da a on he back wall we e no cap u ed due o ou exceeding o he
limi s o he se sea ch dep h o he M3C2 algo i hm. O e all, he applica ions demons a e
possibly achie able accu acies o up o 5 cm, conside ing he pe cen age o de ia ions
wi hin his ange is 69% o he 3D Scanne app, 77% o PolyCam, 83% o Si eScape
and 70% o Scani e se. The p oblem seems o be in he a eas wi h spli ing o une en
su aces due o he d i e o ha accumula es o e ime, showing ha he e is oom o
imp o emen in he so wa e componen o he applica ions.
The e alua ion o he p esen ed esul s is also compa ed wi h he accu acy le els
(LOA) de ined by he U.S. Ins i u e o Building Documen a ion [
50
], widely adhe ed o in
Geoma ics 2023,3571
Scan2BIM p ojec s, and ou lined in Table 3. These LOA le els a e speci ied a he 95 pe cen
con idence le el (2
σ
), a common p ac ice in su eying, e.g., he Ge man s anda d DIN
18710. LOA50 ep esen s he highes class wi h accu acies o up o 1 mm, while LOA10 is
he lowes , indica ing accu acies g ea e han 5 cm. Upon compa ing he alues in Table 2
wi h hose in Table 3, i is e iden ha no applica ion achie es a leas 95% o all dis ances
wi hin he gi en LOA le els up o 5 cm. The achie able accu acies o each applica ion
a e in he ange o 10–20 cm o he 95% con idence le el, signi ying ha he so wa e
componen has no ye achie ed he capabili y o p oduce a highly accu a e poin cloud
ha aligns wi h widely e e enced s anda ds.
Table 2. Nume ical summa y o he cloud- o-cloud compa ison. (S d: s anda d de ia ion).
3D Scanne App PolyCam Si eScape Scani e se
<5 mm 17% 19% 8% 10%
5 mm–1 cm 11% 17% 10% 10%
1–3 cm 30% 32% 46% 31%
3–5 cm 11% 9% 19% 19%
5–10 cm 12% 9% 9% 18%
10–20 cm 11% 12% 6% 9%
20–40 cm 8% 2% 2% 3%
S d (cm) 9 7 6 8
Table 3.
LOA de ini ions (based on de ia ions o 2
σ
) by he U.S. Ins i u e o Building Documen a ion.
Le el Uppe Range Lowe Range
LOA10 Use -de ined 5 cm
LOA20 5 cm 15 mm
LOA30 15 mm 5 mm
LOA40 5 mm 1 mm
LOA50 1 mm 0
The subsequen analysis p io i ized he noise assessmen on he poin clouds on la
su aces, namely walls, loo , and ceiling. To achie e his, clu e , such as u ni u e o wall
accesso ies, co e ing he la su aces was segmen ed away, lea ing behind he ele an
a eas. A plane was i ed in o hese emaining pa s o ep esen he la a eas accu a ely.
The planes we e cons uc ed h ough he andom sampling and consensus (RANSAC)
algo i hm ha calcula es he pa ame e s equi ed o cons uc a co esponding p imi i e
u ilizing a minimum se o poin s [
51
]. The dis ance be ween each poin and he i ed
plane was calcula ed o measu e he noise p esen in he da a se s. This in o ma ion was
isualized o gain insigh s in o he noise le els ac oss he la su aces.
Figu e 5illus a es he esul s o he TLS da a. I is e iden ha he loo and walls
exhibi a smoo h, la su ace, while he ceiling de ia es pa ially om a la su ace, wi h
a ia ions o up o 4 cm along he middle line.
Figu e 6illus a es he dis ances om each poin o he i ed plane along he long
conc e e wall, while Figu e 7 ocuses on he loo and ceiling. A comp ehensi e summa y
o hese compa isons can be ound in Table 4. No ably, no consis en pa e n is obse ed
ac oss all applica ions conce ning hei beha io on each o hese su aces. Fo ins ance,
he 3D Scanne app exhibi s he leas de ia ion on he ceiling su ace, while PolyCam and
Scani e se pe o m be e on he wall su ace, and Si eScape pe o ms be e on he loo
su ace. Wi hin he applica ions, he pe cen age o poin s exceeding a dis ance o 10 cm
emains below 10%, excep o Scani e se. The s anda d de ia ion anges be ween 2 o
7 cm o all applica ions. PolyCam and Si eScape gene ally pe o m be e han he 3D
Scanne app and Scani e se by demons a ing highe poin densi ies in he lowe de ia ion
anges. The de ia ion pa e n iden i ied on he ceiling in he TLS da a (Figu e 5) is no
clea ly e lec ed in he esul s om he applica ions. An impo an ac o con ibu ing o
Geoma ics 2023,3578
7.
S e nbe g, H. De o ma ion Measu emen s a His o ical Buildings wi h Te es ial Lase scanne s. In P oceedings o he ISPRS
Commission V Symposium Image Enginee ing and Vision Me ology, D esden, Ge many, 25–27 Sep embe 2006; pp. 303–308.
A ailable online: h ps://www.isp s.o g/p oceedings/xxx i/pa 5/pape /STER_620.pd (accessed on 17 July 2023).
8.
Wang, W.; Zhao, W.; Huang, L.; Vima lund, V.; Wang, Z. Applica ions o e es ial lase scanning o unnels: A e iew. J. T a ic
T ansp. Eng. (Eng. Ed.) 2014,1, 325–337. [C ossRe ]
9.
Mukupa, W.; Robe s, G.W.; Hancock, C.M.; Al-Manasi , K. A e iew o he use o e es ial lase scanning applica ion o change
de ec ion and de o ma ion moni o ing o s uc u es. Su . Re . 2017,49, 99–116. [C ossRe ]
10.
Raza, M. BIM o Exis ing Buildings: A S udy o Te es ial Lase Scanning and Con en ional Measu emen Technique. Mas e ’s
Thesis, Me opolia Uni e si y o Applied Sciences, Helsinki, Finland, 2017.
11.
Pa k, H.; Lim, S.; T inde , J.; Tu ne , R. 3D su ace econs uc ion o Te es ial Lase Scanne da a o o es y. In P oceedings o
he IGARSS 2010–2010 IEEE In e na ional Geoscience and Remo e Sensing Symposium, Honolulu, HI, USA, 25–30 July 2010;
pp. 4366–4369.
12.
Pe ie, G. An In oduc ion o he Technology Mobile Mapping Sys ems. GeoIn o ma ics
2016
,13, 32–43. A ailable online:
h p://pe ie ied.in o/Pe ie_Mobile_Mapping_Sys ems_Jan-Feb_2010.pd (accessed on 17 July 2023).
13.
Ham az, H.; Con e as, M.A.; Zhang, J. Fo es unde s o y ees can be segmen ed accu a ely wi hin su icien ly dense ai bo ne
lase scanning poin clouds. Sci. Rep. 2017,7, 6770. [C ossRe ]
14.
Chen, D.; Wang, R.; Pee hamba an, J. Topologically Awa e Building Roo op Recons uc ion F om Ai bo ne Lase Scanning Poin
Clouds. IEEE T ans. Geosci. Remo e Sens. 2017,55, 7032–7052. [C ossRe ]
15.
To h, C.; G ejne -B zezinska, D. Rede ining he Pa adigm o Mode n Mobile Mapping. Pho og amm. Eng. Remo e Sens.
2004
,70,
685–694. [C ossRe ]
16.
B iese, C.; Zach, G.; Ve hoe en, G.; Ressl, C.; Ull ich, A.; S udnicka, N.; Doneus, M. Analysis o mobile lase scanning da a and
mul i- iew image econs uc ion. ISPRS-In . A ch. Pho og amm. Remo e Sens. Spa . In . Sci.
2012
,XXXIX-B5, 163–168. [C ossRe ]
17.
S ojano ic, V.; Shoush a i, H.; Aska , C.; Scheide , A.; Schuld , C.; Hellweg, N.; S e nbe g, H. A Concep ual Digi al Twin o 5G
Indoo Na iga ion. 2021. A ailable online: h ps://www. esea chga e.ne /publica ion/351234064 (accessed on 30 No embe
2023).
18.
Ib ahimkhil, M.H.; Shen, X.; Ba a i, K.; Wang, C.C. Dynamic P og ess Moni o ing o Mason y Cons uc ion h ough Mobile
SLAM Mapping and As-Buil Modeling. Buildings 2023,13, 930. [C ossRe ]
19.
Mahdjoubi, L.; Moobela, C.; Laing, R. P o iding eal-es a e se ices h ough he in eg a ion o 3D lase scanning and building
in o ma ion modelling. Compu . Ind. 2013,64, 1272–1281. [C ossRe ]
20.
Sg enza oli, M.; Ba ien os, J.O.; Vassena, G.; Sanchez, A.; Ci ibini, A.; Ven u a, S.M.; Comai, S. Indoo mobile mapping sys ems
and (bim) digi al models o cons uc ion p og ess moni o ing. ISPRS-In . A ch. Pho og amm. Remo e Sens. Spa . In . Sci.
2022
,
XLIII-B1-2, 121–127. [C ossRe ]
21.
Leh ola, V.V.; Nikoohema , S.; Nüch e , A.; Leh ola, V.V.; Nikoohema , S.; Nüch e , A. Indoo 3D: O e iew on Scanning and
Recons uc ion Me hods. In Handbook o Big Geospa ial Da a; We ne , M., Chiang, Y.-Y., Eds.; Sp inge : Be lin, Ge many, 2021;
pp. 55–97. [C ossRe ]
22.
Lacha , E.; Mache , H.; Mi e , M.-A.; Landes, T.; G ussenmeye , P. Fi s expe iences wi h kinec 2 senso o close ange 3d
modelling. ISPRS-In . A ch. Pho og amm. Remo e Sens. Spa . In . Sci. 2015,XL-5/W4, 93–100. [C ossRe ]
23.
Khoshelham, K.; T an, H.; Acha ya, D. Indoo mapping eyewea : Geome ic e alua ion o spa ial mapping capabili y o hololens.
ISPRS-In . A ch. Pho og amm. Remo e Sens. Spa . In . Sci. 2019,XLII-2/W13, 805–810. [C ossRe ]
24. Delasse, C.; La ki i, H.; Hajji, R.; Rached, I.; Landes, T. Indoo 3D Recons uc ion o Buildings ia Azu e Kinec RGB-D Came a.
Senso s 2022,22, 9222. [C ossRe ]
25.
Hübne , P.; Clin wo h, K.; Liu, Q.; Weinmann, M.; Wu s ho n, S. E alua ion o HoloLens T acking and Dep h Sensing o Indoo
Mapping Applica ions. Senso s 2020,20, 1021. [C ossRe ]
26.
Weinmann, M.; Wu s ho n, S.; Ju zi, B. Semi-au oma ic image-based co- egis a ion o ange imaging da a wi h di e en
cha ac e is ics. ISPRS-In . A ch. Pho og amm. Remo e Sens. Spa . In . Sci. 2013,XXXVIII-3, 119–124. [C ossRe ]
27. Kalan a i, M.; Nechi o , M. 3D Indoo Su eying—A Low Cos App oach. Su . Re . 2016,49, 1–6. [C ossRe ]
28.
Leh ola, V.V.; Kaa inen, H.; Nüch e , A.; Kaijaluo o, R.; Kukko, A.; Li key, P.; Honka aa a, E.; Rosnell, T.; Vaaja, M.T.; Vi anen,
J.-P.; e al. Compa ison o he Selec ed S a e-O -The-A 3D Indoo Scanning and Poin Cloud Gene a ion Me hods. Remo e Sens.
2017,9, 796. [C ossRe ]
29.
Tucci, G.; Visin ini, D.; Bono a, V.; Pa isi, E.I. Examina ion o Indoo Mobile Mapping Sys ems in a Di e si ied In e nal/Ex e nal
Tes Field. Appl. Sci. 2018,8, 401. [C ossRe ]
30.
di Filippo, A.; Sánchez-Apa icio, L.J.; Ba ba, S.; Ma ín-Jiménez, J.A.; Mo a, R.; Aguile a, D.G. Use o a Wea able Mobile Lase
Sys em in Seamless Indoo 3D Mapping o a Complex His o ical Si e. Remo e Sens. 2018,10, 1897. [C ossRe ]
31.
Salgues, H.; Mache , H.; Landes, T. E alua ion o mobile mapping sys ems o indoo su eys. ISPRS-In . A ch. Pho og amm.
Remo e Sens. Spa . In . Sci. 2020,XLIV-4/W1-2020, 119–125. [C ossRe ]
32.
Wilkening, J.; Kapaj, A.; C on, J. C ea ing a 3D Campus Rou ing In o ma ion Sys em wi h A cGIS Indoo s. In D eilände agung de
OVG, DGPF und SGPF Pho og amme ie-Fe ne kundung-Geoin o ma ion-2019; Thomas Ke s en: Hambu g, Ge many, 2019.
33.
Mu iyoso, A.; G ussenmeye , P.; Landes, T.; Mache , H. Fi s assessmen s in o he use o comme cial-g ade solid s a e lida o
low cos he i age documen a ion. ISPRS-In . A ch. Pho og amm. Remo e Sens. Spa . In . Sci.
2021
,XLIII-B2-2, 599–604. [C ossRe ]

Geoma ics 2023,3579
34.
Shoush a i, H.; Willemsen, T.; S e nbe g, H. Many Ways Lead o he Goal—Possibili ies o Au onomous and In as uc u e-Based
Indoo Posi ioning. Elec onics 2021,10, 397. [C ossRe ]
35.
Tanskanen, P.; Kole , K.; Meie , L.; Camposeco, F.; Sau e , O.; Polle eys, M. Li e Me ic 3D Recons uc ion on Mobile Phones.
In P oceedings o he 2013 IEEE In e na ional Con e ence on Compu e Vision (ICCV), Sydney, Aus alia, 1–8 Decembe 2013;
pp. 65–72.
36.
Ke s en, T.P. The Sma phone as a P o essional Mapping Tool | GIM In e na ional. GIM In e na ional, 25 Feb ua y 2020. A ailable
online: h ps://www.gim-in e na ional.com/con en /a icle/ he-sma phone-as-a-p o essional-mapping- ool (accessed on 17
July 2023).
37.
Wikipedia. Tango (Pla o m)-Wikipedia. A ailable online: h ps://en.wikipedia.o g/wiki/Tango_(pla o m) (accessed on 10
May 2023).
38.
Bianchini, C.; Ca ena, L. The Democ a iza ion o 3D Cap u ing an Applica ion In es iga ing Google Tang Po en ials. In . J. Bus.
Hum. Soc. Sci. 2019,12, 3298576. [C ossRe ]
39.
Google. Build New Augmen ed Reali y Expe iences ha Seamlessly Blend he Digi al and Physical Wo lds | ARCo e | Google
De elope s. A ailable online: h ps://de elope s.google.com/a (accessed on 10 May 2023).
40.
Diaki é, A.A.; Zla ano a, S. Fi s expe imen s wi h he ango able o indoo scanning. ISPRS Ann. Pho og amm. Remo e Sens.
Spa . In . Sci. 2016,III-4, 67–72. [C ossRe ]
41.
F oehlich, M.; Azha , S.; Van u e, M. An In es iga ion o Google Tango
®
Table o Low Cos 3D Scanning. In P oceedings o he
34 h In e na ional Symposium on Au oma ion and Robo ics in Cons uc ion, Taipei, Taiwan, 28 June–1 July 2017; pp. 864–871.
42.
Boboc, R.G.; Gî bacia, F.; Pos elnicu, C.C.; Gî bacia, T. E alua ion o Using Mobile De ices o 3D Recons uc ion o Cul u al
He i age A i ac s. In VR Technologies in Cul u al He i age; Duguleană, M., Ca ozzino, M., Gams, M., Tanea, I., Eds.; Communica-
ions in Compu e and In o ma ion Science; Sp inge In e na ional Publishing: Cham, Swiz e land, 2019; Volume 904, pp. 46–59.
[C ossRe ]
43.
Lue zenbu g, G.; K oon, A.; Bjø k, A.A. E alua ion o he Apple iPhone 12 P o LiDAR o an Applica ion in Geosciences. Sci. Rep.
2021,11, 1–9. [C ossRe ]
44.
Riquelme, A.; Tomás, R.; Cano, M.; Pas o , J.L.; Jo dá-Bo deho e, L. Ex ac ion o discon inui y se s o ocky slopes using iPhone-12
de i ed 3DPC and compa ison o TLS and S M da ase s. IOP Con . Se . Ea h En i on. Sci. 2021,833, 012056. [C ossRe ]
45.
Losè, L.T.; Sp ea ico, A.; Chiab ando, F.; Tonolo, F.G. Apple LiDAR Senso o 3D Su eying: Tes s and Resul s in he Cul u al
He i age Domain. Remo e Sens. 2022,14, 4157. [C ossRe ]
46.
Díaz-Vila iño, L.; T an, H.; F ías, E.; Balado, J.; Khoshelham, K. 3D mapping o indoo and ou doo en i onmen s using Apple
sma de ices. ISPRS-In . A ch. Pho og amm. Remo e Sens. Spa . In . Sci. 2022,XLIII-B4-2, 303–308. [C ossRe ]
47.
Sp ea ico, A.; Chiab ando, F.; Losè, L.T.; Tonolo, F.G. The ipad p o buil -in lida senso : 3D apid mapping es s and quali y
assessmen . ISPRS-In . A ch. Pho og amm. Remo e Sens. Spa . In . Sci. 2021,XLIII-B1-2, 63–69. [C ossRe ]
48.
Zolle +F öhlich. Z+F IMAGER
®
5016: Zolle +F öhlich. A ailable online: h ps://www.zo e.de/lase scanne /3d-lase scanne /
z- -image -5016 (accessed on 17 July 2023).
49.
CloudCompa e. ( e sion. 2.12.4) [GPL So wa e]. A ailable online: h p://www.cloudcompa e.o g/ (accessed on 14 July 2022).
50.
U.S. Ins i u e o Building Documen a ion. USIBD Le el o Accu acy (LOA) Speci ica ion Guide, 2.0-2016. 2016. A ailable
online: h ps://cdn.ymaws.com/www.nysapls.o g/ esou ce/ esmg /2019_con e ence/handou s/hale-g_bim_loa_guide_c1
20_ 2.pd (accessed on 12 Oc obe 2022).
51.
Schnabel, R.; Wahl, R.; Klein, R. E icien RANSAC o poin -cloud shape de ec ion. In P oceedings o he 2007 Compu e
G aphics Fo um, Honolulu, HI, USA, 25–29 June 2017; pp. 214–226. [C ossRe ]
Disclaime /Publishe ’s No e:
The s a emen s, opinions and da a con ained in all publica ions a e solely hose o he indi idual
au ho (s) and con ibu o (s) and no o MDPI and/o he edi o (s). MDPI and/o he edi o (s) disclaim esponsibili y o any inju y o
people o p ope y esul ing om any ideas, me hods, ins uc ions o p oduc s e e ed o in he con en .