International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B1,2012 
XXII ISPRS Congress, 25 August - 01 September 2012, Melbourne, Australia 
THE PERFORMANCE OF A TIGHT INS/GNSS/PHOTOGRAMMETRIC INTEGRATION 
SCHEME FOR LAND BASED MMS APPLICATIONS IN GNSS DENIED 
ENVIRONMENTS 
Chien-Hsun Chu a ' *, Kai-Wci Chiang a , 
Dept, ol Geomatics, National Cheng Kung University, Taiwan -chienhsun0229@msn.com 
Commission IAG and I C I: New I rends in Direct Geo-referencing Technologies 
KEY WORDS: Direct georeferencing, LC, TC, MMS 
ABSTRACT: 
Nowadays the most common technologies used for positioning and orientation of a mobile mapping system include using Global 
Navigation Satellite System (GNSS) as a major positioning sensor and Inertial Navigation System (INS) as the major orientation 
sensor. The integration strategy ot the most commercially system is the loosely coupled (LC) architecture, that has the simplest 
architecture using the GNSS solutions to aid the INS navigation information with proper optimization estimator. The LC does 
combine the two sensors’ solutions when the number of tracked satellite is more than four. In recent year, another commonly 
integration strategy is known as tightly coupled (TC) architecture. Because the TC uses the GNSS measurements to aid INS, it can 
integrate measurements provided by GNSS receiver and INS unless no GNSS satellite is tracked. Obviously, the TC architecture is a 
better candidate for land based mobile mapping applications than LC in Taiwan. Unfortunately, there are still many GNSS denied 
environment in the urban area, therefore the TC architecture is still not robust and stable enough for MMS application. The overall 
objective of this paper is to provide a scheme that tightly integrates INS/GNSS and Photogrammetric for land based MMS 
applications with sufficient and stable POS solutions during GNSS outages. In the traditional photogrammetry operation, numerous 
ground control points are applied to compute those Exterior Orientation Parameters (EOPs) of cameras by bundle adjustment. The 
key opinion is to derive the INS centre position and attitude and reconstruct 3-D tracking and 3-D object space by cameras EOPs. 
The proposed algorithm is verified using field test data collected in GNSS denied environments and the preliminary results presented 
in this study illustrated that the proposed algorithm is able to provide 60% improvement in terms of positioning and orientation 
accuracy in Taipei and Tainan cities. 
1 INTRODUCTION 
The development of land-based mobile mapping systems was 
initiated by two research groups, namely The Center for 
Mapping at Ohio State University, USA, and the Department of 
Geomatics Engineering at the University of Calgary, Canada. In 
the early 2000s, a number of land-based mobile mapping 
systems have been utilized in commercial applications (Chiang 
et al., 2008). The process of mobile mapping involves 
producing multiple images of a given object from various 
positions. The three-dimension (3D) positions of the object with 
respect to the camera frame can then be measured. An Inertial 
Navigation System (INS) is a self-contained navigation 
technique in which measurements provided by accelerometers 
and gyroscopes are used to track the position and orientation of 
an object relative to a known starting point, orientation and 
velocity. The Global Navigation Satellite Systems (GNSS) is a 
universal, all-weather, world-wide positioning system that 
provides time, position, and velocity data. Both systems can be 
used as stand-alone navigation tools or in conjunction with 
other sensors for various purposes. Moreover, the integration of 
GNSS and INS can overcome problems with environments such 
as urban canyons, forests, and indoor settings, where GNSS 
alone cannot provide service. In order to attain reasonable 
accuracies of position and orientation solutions, a tactical grade 
or higher quality INS along with GNSS has been applied as the 
primary position and orientation system for current commercial 
systems. Unfortunately, the cost of such systems is still at such a 
high level thus the popularity of mobile mapping systems 
remains limited, especially due to the price of the Inertial 
Measurement Unit (IMU). However, advances in MEMS 
technology enables the development of complete IMU 
composed of multiple MEMS-based accelerometers and 
gyroscopes. In addition to their compact and portable size, the 
price of MEMS-based systems is far less than those of high 
quality IMU. 
In the classical approach, the KF is applied in real-time 
applications to fuse different data from various sensors while 
optimal smoothing is applied in the post-mission mode. The 
basic idea of using the KT in INS/GNSS integration is to fuse 
independent and redundant sources of navigation information 
with a reference navigation solution to obtain an optimal 
estimate of navigation states, such as position, velocity and 
orientation. Furthermore, for INS/GNSS integration 
applications where the process and measurement models are 
nonlinear, the Extended KF (EKF) also work for nonlinear 
dynamic systems with a non-Gaussian distribution, except for 
heavily skewed nonlinear dynamic systems, where EKF may 
experience problems. The EKF simply applies the first order 
term of the Taylor series expansion for the approximation of a 
nonlinear system and the probability density function is 
Corresponding author: C.H. Chu