Using Relative Orientation to Improve the
Accuracy of Exterior Orientation Parameters
of Low Cost POS
Wenlin Yan, Luísa Bastos, Sérgio Madeira, Américo Magalhães, José A. Gonçalves
Abstract
A simple and inexpensive
POS
for application to light airborne
platforms was developed, and a modified Kalman Filtering
was designed to integrate
GNSS/IMU/
Image data, which uses the
relative orientation information to improve the accuracy of the
exterior orientation parameters. The precise relative orientation
of conjugate images was obtained using a
SIFT/SFM
matching
algorithm. The relative exterior orientations were transformed
from the camera frame to the navigation frame before they were
used as external update information of the Kalman Filter. Com-
bining all the relative orientation information retrieved from
images, the Kalman Filter can give an improved output for the
exterior orientation parameters. Airborne result from the tests
of one straight strip shows that the heading accuracy of a
GNSS/
MEMS-IMU
was improved from 0.26° to 0.10°, and the result from
one closed strip shows that heading accuracy was improved
from 1.11° to 0.85° and roll accuracy from 0.54° to 0.43°.
Introduction
The integration of and
IMU
(Inertial Measurement Unit) and
GNSS
(Global Navigation Satellite System), also named as
POS
(Position and Orientation System), can directly provide pre-
cise exterior parameters for the application of Direct Geo-ref-
erencing (
DG
) with none or few
GCPs
(Ground Control Points).
This has great advantages, especially in terms of cost and
efficiency, in comparison with conventional photogrammetry
methodologies. The
POS
is becoming a very popular method
for the survey works, for example in urban street mapping
(Madeira
et al.
, 2010), sandy beaches or shoreline changes
monitoring (Madeira
et al.
, 2013), forests monitoring (Hopkin-
son
et al.
, 2013), and particularly in quick mapping of disaster
areas (Hutton and Mostafa, 2005; Mitishita
et al.
, 2008) where
methods that allow a fast characterization of the actual situa-
tion are of utmost relevance.
Some commercial
POS
, such as Applanix
POS/AV
from Trim-
ble Company and AEROcontrol from IGI company, which are
using tactical or navigation grade
IMU
s, integrated with high
quality cameras/scanners, have the ability to provide prod-
ucts with high level of accuracy, but with the disadvantage
of being very expensive systems. These high grade
POS
can
be mounted in stable platforms, such as terrestrial motorcars
or standard airplanes, which have enough capability to carry
these dedicated devices. The low cost
POS
that integrates
GNSS
and
MEMS
-
IMU
(
MEMS
, Micro Electro Mechanical System),
which is lighter and smaller, is acquiring a growing relevance
due to the potential of its application in remote sensing from
lightweight airborne platforms, including
UAV
s whose market
is flourishing. However, the accuracy of exterior parameters
provided by the low cost
POS
is not enough to obtain precise
photogrammetric products.
The accuracy of the integrated
GNSS/IMU
can be improved
by optimizing the integration strategies, exploiting different
coupling strategies, such as loosely/tightly/deeply, and proper
filtering (Xia
et al.
, 2016), and/or by incorporating information
from multi-antenna systems (Tomé, 2002), terrain aiding (Zhou
et al.
, 2016), Radar odometer (Quist and Beard, 2016) or other
sensors.
Besides the classical integration of
GNSS
and
IMU
s or other
sensors, other strategies using the fusion of
GNSS/IMU
/Imag-
es become an optional method to improve the accuracy in
navigation and photogrammetry. For example, sequential
AT
(Aerial Triangulation) method using high correlated images
can produce accurate results for airborne applications at a
low cost (Choi and Lee, 2013);
SLAM
(Simultaneous Location
and Mapping) method, which was introduced in robotics and
artificial intelligence, gives an approach to locate a position
at an unknown place with the help of the relationship of
the landmarks or image features. This type of approach has
now been implemented in many other application scenarios,
such as indoor positioning, underwater, or aerial navigation
(Soeder and Raquet, 2015). The critical problem of the
SLAM
technique is to estimate the correlations between the different
landmarks (Durrant-Whyte, 1988, Durrant-Whyte and Bailey,
2006), which grow with successive observations; the more
these correlations grow, the better the solution will be. The
SLAM
strategy plays an important role in the applications in
which the
GNSS/IMU
solution is poor (particularly in the case
of
GNSS
missing). Previous research showed that the
SLAM
po-
sition and orientation accuracy can satisfy the demand of the
Wenlin Yan is with the Department of Geosciences Environ-
ment and Spatial Planning, Faculty of Science, University of
Porto, Rua do Campo Alegre 687, 4169-007, Porto, Portugal;
and with the Institute of Space Sciences, Shandong Universi-
ty, 180 Rd. Wenhuaxilu, 264209, Weihai, Shandong, China.
Luísa Bastos and José A. Gonçalves are with the Department
of Geosciences Environment and Spatial Planning, Faculty of
Science, University of Porto, 687 Rua do Campo Alegre, 4169-
007, Porto, Portugal; and with CIIMAR, University of Porto,
Terminal de Cruzeiros do Porto de Leixões, Avenida General
Norton de Matos, 4450-208, Matosinhos, Portugal.
Américo Magalhães is with the Astronomical Observatory of
the Faculty of Science, University of Porto, Monte da Virgem,
4430-146 Vila Nova de Gaia, Portugal.
Sérgio Madeira (
) is with the University of
Trás-os-Montes e Alto Douro, 5000-801, Vila Real, Portugal;
and with INESC TEC - Technology and Science, Rua Dr. Ro-
berto Frias, 4200-465, Porto, Portugal.
Photogrammetric Engineering & Remote Sensing
Vol. 83, No. 2, February 2017, pp. 153–161.
0099-1112/17/153–161
© 2017 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.83.2.153
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
February 2017
153
