INTEGRATION OF LASER AND PHOTOGRAMMETRIC DATA FOR CALIBRATION 
PURPOSES : 
A. F. Habib *, M. S. Ghanma?, M. F. Morgan“, E. Mitishita” 
* Department of Geomatics Engineering, University of Calgary 
2500, University Drive NW, Calgary AB T2N 1N4 Canada — (habib, mghanma)@geomatics.ucalgary.ca, mfmorgan(ucalgary.ca, 
b Departamento de Geomática, Universidade Federal Do Paraná, Caixa Postal 19.001, 81.531-970 Curitiba, Parana, Brasil 
mitishita@ufpr.br 
TS — PS: WG V5 Platform and Sensor Integration 
KEY WORDS: Laser scanning, Registration, Calibration, Fusion, Feature, Modelling 
ABSTRACT: 
Laser scanners are becoming popular since they provide fast and dense geometric surface information. However, sudden elevation 
changes along the surface are not clearly visible in the laser data due to the sparse distribution of captured points. In general, laser 
data provides high density surface information in homogenous areas and low density surface information elsewhere (i.e., object 
space break-lines). Photogrammetry, on the other hand, provides less dense surface information but with high quality, especially 
along object space discontinuities. Hence, a natural synergy of both systems can be inferred and consequently integration of the 
respective data would lead to higher quality surface information than that obtained through the use of a single sensor. However, prior 
to such integration, both systems should be precisely calibrated and aligned. The calibration is usually carried-out for each system 
independently using additional control information. In this paper, the calibration of the laser and photogrammetric systems is 
evaluated by checking the quality of fit between co-registered photogrammetric and laser surfaces. The paper starts by introducing a 
registration procedure where a set of linear features is extracted from both sets. First, planar surfaces from laser data are extracted 
and adjacent planes are intersected to determine three-dimensional straight line segments. Secondly, linear features from the 
photogrammetric dataset are obtained through aerial triangulation. A mathematical model for expressing the necessary constraints 
for the alignment of conjugate photogrammetric and laser straight lines is established. The model ensures that corresponding straight 
lines will be collinear after registering the two datasets relative to a common reference frame. The quality of fit between the 
registered surfaces is then used to evaluate and/or improve the calibration parameters of the photogrammetric and laser systems. In 
this paper, an experiment with real data is used to illustrate this concept. The registered surfaces in this example revealed the 
presence of systematic inconsistencies between the photogrammetric and laser systems. The pattern of these inconsistencies is found 
to resembie the effect of un-calibrated lens distortion. In this case, the laser data is used as control information for the determination 
of lens distortion, which when considered leads to a better fit between the registered surfaces. The estimated lens distortion using the 
laser was found to be very close to that determined through a rigorous camera calibration procedure. 
1. INTRODUCTION populated areas. Richness in semantic information and dense 
positional information along object space break lines add to its 
Laser scanners are becoming an increasingly accepted tool for advantages. Nonetheless, photogrammetry has its own 
acquiring 3D point clouds that represent scanned objects with drawbacks; where there is almost no positional information 
millimetre precision. As can be inferred from the name, laser along homogeneous surfaces and vertical accuracy is worse 
scanning is a non-contact range measurement based on emitting than the planimetric accuracy. A major existing obstacle in the 
a laser light pulse and instantaneously detecting the reflected way of automation in photogrammetry is the complicated and 
signal. This should be coupled with high-quality GPS/INS units sometimes unreliable matching procedures, especially when 
for tracking the position and orientation of the range finder as it dealing with large scale imagery over urban areas. 
Scans over objects and sugfaces under consideration, It can be clearly observed that both, photogrammetry and laser 
The sparse and positional nature of laser data makes it ideal for data, have unique characteristics that make them preferable in 
mapping homogeneous surfaces but lacks the ability to capture certain applications. One can notice that a disadvantage in one 
objects’ break-lines with reliable quality. Another drawback is technology is contrasted by an opposite strength in the other. 
that laser data has no mherent redundancy and its planimetric Hence, integrating the two systems would lead to higher quality 
accuracy is worse than the vertical (Maas, H.-G., 2002), in surface information — (Baltsavias, 1999). However, the 
addition to little or no semantic information. However, the complementary information can be fully utilized only alter 
continuous development of laser systems, in the aspect of precise calibration of both systems, which is separately 
reduced hardware size and increased resolution and density, implemented for each system. The synergy would be 
makes it an increasingly favoured option in a variety of considered complete after aligning the photogrammetric and 
applications especially where rapid and accurate data collection laser data models relative to a common reference frame. (Habib 
on physical surface is required (Schenk and Csathó, 2002). and Schenk, 1999; Postolov et al., 1999). 
On the other side, photogrammetric data is characterized by This paper introduces a registration procedure through which 
high redundancy through observing the desired features in the calibration of photogrammetric and laser scanning systems 
multiple images making it more suited for mapping heavily is assessed. The suggested technique emphasizes the type and 
170 
[nternai 
extracti 
registra 
discrep: 
Most rc 
primit 
dataset: 
scannec 
instead 
(Baltsa: 
registra 
interpol 
is then 
the ele 
1994; I 
approac 
especia 
areas. 
surface: 
horizon 
et al. (| 
original 
implem 
surface 
Additio 
differen 
Schenk 
distance 
to loc: 
minimi 
implem 
automat 
process 
planar 
photogr 
object 
informa 
In this | 
primitiv 
photogr 
frame. 
general 
each cc 
under 
experim 
recomm 
The reg 
acquirec 
and enh 
attained 
subsecti 
effectivi 
2.1 Re 
To regis 
be ident 
subsequ 
datasets 
influenc 
first dec 
transfori