Aiding Indoor Photogrammetry with UWB Sensors
Andrea Masiero, Francesca Fissore, Alberto Guarnieri, Francesco Pirotti, and Antonio Vettore
Abstract
Given the worldwide spread of smartphones, photogram-
metric surveying with mobile devices is becoming of signifi-
cant interest in the research community for providing low
cost three-dimensional (
3D
) models. Since the photogram-
metric procedure applied to images produces a projective
model, some external information is needed in order to
obtain a 3D metric model. To th
Satellite System (
GNSS
) measure
to the photogrammetric reconst
ever, the quality of the obtained reconstruction is related to
the
GNSS
positioning accuracy, which is typically at meter-
level for cheap receivers as those embedded in most of the
consumer mobile devices, e.g. smartphones. Furthermore,
this approach cannot be used in
GNSS
-denied environ-
ments (e.g. indoors). To overcome these issues, this paper
investigates the integration of information provided by an
Ultra-Wideband (
UWB
) positioning system with image-based
reconstruction to produce a metric reconstruction. Further-
more, the orientation (with respect to North-East directions)
of the model is assessed thanks to the use of inertial sen-
sors included in the
UWB
devices. Results of this integration
are shown on two case studies in indoor environments.
Introduction
The worldwide spread of applications based on spatial data
and position information (e.g. location-based services) is
causing a continuously increasing quest for three-dimensional
(
3D
) data and motivating the development of procedures and
devices in order to ease their generation in any kind of work-
ing condition.
Among the available techniques for
3D
data generation,
photogrammetry is one of the most commonly used thanks
to its relatively low cost (nowadays standard cameras are fre-
quently used to generate quite good models) and to the easi-
ness of usage of photogrammetric software currently available
on the market (Gonzalez-Aguilera
et al.
2018), in particular
those based on the implementation of the Structure from Mo-
tion approach.
When dealing only with images, the photogrammetric
procedure produces just projective reconstructions, hence the
generation of
3D
metric models (e.g. properly scaled) requires
the introduction of certain external information: control
points and Global Navigation Satellite System (
GNSS
) mea-
surements are often used for such purpose (where the latter
can be used for georeferencing the
3D
model as well). Despite
the introduction of
GNSS
measurements in the photogram-
metric reconstruction procedure, which is very common and
usually leads to reliable and accurate results, it is clearly a
nonviable way in
GNSS
-denied environments, such as indoors
(Dabove, Di Pietra, and Lingua 2018; Tucci
et al.
2018).
Actually, the growing interest in surveying areas during/
just after natural disasters, e.g. difficult to reach with terres-
trial vehicles, is also motivating the development of low-cost
direct georeferencing techniques (Chiang, Tsai, and Chu 2012;
Lo
et al.
2015; Bendea
et al.
2008; Pfeifer, Glira, and Briese
2012), i.e. where reconstruction is obtained by exploiting high
grade positioning sensors mounted on the vehicle (e.g.
GNSS
and inertial sensors), without using control points.
The integration of information provided by navigation/
inertial sensors in the photogrammetric reconstruction pro-
cedure has also been recently considered in the framework
of mapping with portable mobile devices, e.g. smartphones
iesi
et al.
2017; Masiero
et al.
2016), where such informa-
allows to obtain metric reconstructions in
GNSS
-denied
ironments as well (Mustaniemi
et al.
2017; Ham, Lucey,
and Singh 2014; Alsubaie, Youssef, and El-Sheimy 2017).
Motivated by the above considerations, this paper consid-
ers the integration of information provided by Ultra-Wide-
band (
UWB
) sensors in the photogrammetric reconstruction
procedure in order to obtain
3D
metric models. Interestingly,
the use of
UWB
sensors can be simultaneously used also for
navigation purposes.
UWB
sensors are radio transmitters/receivers which enable
real-time position estimation of an
UWB
rover. Such position
is obtained by means of trilateration, by combining infor-
mation of range measurements collected from a set of
UWB
devices, named anchors, fixed at constant locations. A proper
UWB
sensor calibration can be considered in order to im-
prove positioning performance, e.g. systematic error has been
modeled as a constant time lag (due to device synchroniza-
tion) (Hol 2011) and as a polynomial function of the distance
(Dierenbach
et al.
2015; Toth
et al.
2015). Since positioning is
based on the use of a collection of range measurements, best
accuracy of the method is achieved when all measurements
are collected in clear line of sight (
CLOS
), i.e. when the line
connecting two
UWB
devices is not obstructed by any obstacle.
Unfortunately, multipath and nonclear line of sight measure-
ments typically cannot be completely avoided, hence they
usually affect the positioning accuracy.
Similar to (Masiero
et al.
2018), in this work an
UWB
rover
was rigidly attached to a standard camera: their relative posi-
tion and orientation were constant during all image acquisi-
tion. Consequently, metric reconstruction can be achieved
by estimating the scale factor comparing the camera-rover
positions provided by the
UWB
system with the corresponding
locations in the photogrammetric reconstruction. Since the
considered rover is also provided with an Inertial Measure-
ment Unit (
IMU
), its measurements can be used also for orient-
ing the
3D
model according to the East-North directions.
The main advantages of this approach are:
• Wide range of working conditions. It represents an easy and
low-cost method for obtaining
3D
photogrammetric recon-
structions in a wide range of working conditions, and, in
particular, in
GNSS
-denied environments.
• Portability. It only requires the usage of easily portable
(small and lightweight) devices.
Interdepartmental Research Center of Geomatics (CIRGEO),
University of Padova, Viale dell’Università 16, Legnaro (PD)
35020, Italy (
).
Photogrammetric Engineering & Remote Sensing
Vol. 85, No. 5, May 2019, pp. 369–378.
0099-1112/18/369–378
© 2019 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.85.5.369
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
May 2019
369
