AN APPROACH TO OPTIMIZE REFERENCE GROUND CONTROL REQUIREMENTS 
FOR ESTIMATING LIDAR/IMU BORESIGHT MISALIGNMENT 
A. Pothou 3 ’*, C. Toth c , S. K.aramitsos b , A. Georgopoulos 3 
a Laboratory of Photogrammetry - (apothou, drag)@central.ntua.gr 
b Laboratory of Higher Geodesy - karamits@central.ntua.gr 
School of Rural & Surveying Engineering, National Technical University of Athens, Greece 
c Center for Mapping, The Ohio State University, 1216 Kinnear Road, Columbus, OH 43212 USA - 
toth@cfm.ohio-state.edu 
WG 1/2 
KEY WORDS: Boresight misalignment, GPS, IMU, Direct Georeferencing, MMS, LiDAR, QA/QC 
ABSTRACT: 
LiDAR systems are complex multi-sensory systems and include at least three main sensors: GPS, IMU navigation sensors, and the 
laser-scanning device. High-performance integrated GPS/IMU systems provide the navigation solution for the LiDAR data 
acquisition platform, and therefore, the proper calibration, including individual and inter-sensor calibration, is a must to achieve the 
highest accuracy of the output data. Specifically regarding the boresight misalignment, the spatial relationship between the IMU 
body frame and the LiDAR body frame is of high importance as it could be the largest source of systematic errors in airborne MMS, 
and thus must be determined before the system can be effectively utilized. In this research, the feasibility of using urban areas for 
boresight misalignment is investigated. In particular, the impact of the building shape, size, distribution, etc. on the performance of 
the boresight misalignment process, is of interest. In this study, photogrammetrically restituted buildings were used as the reference 
surfaces, called ‘building-positions’ or ‘reference-positions’. The influence of the number of ‘building-positions’ and their 
distribution on the boresight’s misalignment parameter estimation is investigated and evaluated through QA/QC statistical tests. 
1. INTRODUCTION 
LiDAR (Light Detection And Ranging, also known as Airborne 
Laser Scanning - ALS) is a highly automated and still rapidly 
evolving technology, with excellent vertical accuracy of point 
measurements. LiDAR has many benefits, and is quickly 
becoming the prime technology for large-scale acquisition of 
elevation data due to its capability to directly measure 3D 
coordinates of a huge number of points. LiDAR systems are 
complex multi-sensory systems including: GPS (Global 
Positioning System), and IMU (Inertial Measurement Unit, also 
known as INS Inertial Navigation System) navigation sensors, 
and the laser-scanning device. Most of the new systems also 
include a medium format digital camera to provide 
conventional image coverage of the surveyed area. A variety of 
highly specialized systems based on modem imaging sensors, 
such as CCD cameras, LiDAR, and hyper/multi-spectral 
scanners, have been developed in the last decade. LiDAR is 
considered as a basic component of airborne Mobile Mapping 
Systems (MMS) (for details see Bossier and Toth, 1995; 
Schwarz et al., 1993; El-Sheimy et al., 1995). The proper 
calibration of this MMS, including individual and inter-sensor 
calibration, is a must to achieve the highest accuracy of the 
output data. Particularly regarding the boresight misalignment, 
the spatial relationship between the IMU body frame and the 
LiDAR body frame is of high importance, as it could be the 
largest source of systematic errors in airborne MMS, and thus, 
must be determined before the system can be effectively 
utilized (Burman, 2000). In most installations, the lever arms 
between LiDAR/GPS/IMU sensors can be determined 
separately by independent means, with good accuracy. In sharp 
contrast, the determination of the boresight angles is only 
possible in-flight once the GPS/IMU derived orientation 
becomes sufficiently accurate (Skaloud and Lichti, 2006). 
Despite several years of progress, the boresight estimation 
between the LiDAR and IMU sensors is still heavily researched. 
Baltsavias (1999) presents an overview of basic relations and 
error formulas concerning airborne laser scanning. Also a large 
number of publications report the existence of systematic errors 
(Schenk, 2001; Filin, 2001). The solutions for dealing with and 
eliminating the effect of systematic boresight misalignment 
errors can be categorized into two groups. Techniques in the 
first group are based on the introduction of a correction 
transformation of the laser points to minimize the difference 
between the corresponding LiDAR points and the ground truth; 
for instance, Kilian et al. (1996), Pothou et al. (2007), use 
surface patches while Csanyi and Toth, (2007) investigate the 
achievable LiDAR data accuracy improvement using LiDAR- 
specific ground control targets. This technique is frequently 
called data driven. In contrast, the other group attempts to 
rigorously model the system to recover the systematic errors 
(Burman, 2000) and treats the discrepancies between 
overlapping strips, including navigation and sensor calibration 
errors, as orientation errors. 
Since the ground surfaces are not always known, or at the 
required accuracy level, preference has been given to 
techniques which do not require a priori knowledge of the 
surface (Toth and Csanyi, 2001; Toth et al., 2002). This 
alternative solution is independent from ground control, and can 
Corresponding author