ANALYSIS AND ACCURACY ASSESSMENT OF AIRBORNE LASERSCANNING
SYSTEM
Abdullatif Alharthy', James Bethel?, Edward M. Mikhail?
'College of Engineering, Umm Al-Qura University, P.O. Box 555, Makkah, Saudi Arabia
2 School of Civil Engineering, Purdue University, 550 Stadium Mall Drive, West Lafayette, IN 47907
alharthy@uqu.edu.sa, bethel@ecn.purdue.edu, mikhail@ecn.purdue.edu
KEY WORDS: LIDAR, Adjustment, Calibration, Prediction, Error, Accuracy, Performance, Quality
ABSTRACT:
Airborne laser scanning technology is impressive in its capability of collecting a tremendous number of points in a very short time
and providing a reasonable depiction of complex objects in the scanned areas. So far it has been used in a wide range of applications
with promising results. Since it is in a very early stage of development, users are still trying to determine the best ways to collect
and analyze the data. The quality of any final product naturally depends on the original data and methods of generating it. Thus the
quality of the data should be verified before assessing any of its products. The work described in this paper is aimed at a quantitative
accuracy evaluation of the laser data itself. This is an area that has been under-emphasized in much published work on the
applications of airborne laser scanning data. The evaluation is done by field surveying, including triangulation and leveling. The
results will address both the planimetric as well as the height accuracy of the laser data.
1. INTRODUCTION
With the recent increase in the scope of laser altimetry
applications, there is a need for more studies to be conducted
on the data quality assessment and on means of improving data
quality. Generally, as documented in many LIDAR system
vendors' specifications, the accuracy of individual data points
is about 5-15cm in height and about 30-50cm in planimetry.
However, those values might be degraded if the data collection
is carried out in less than ideal conditions (Baltsavias, 1999).
This paper outlines the work that has been done to assess and
quantify the quality of the laser scanning data that was
collected over Purdue University campus in spring 2001 and
used in this research. As an introduction, major sources of
errors in laser data are briefly discussed. Then a detailed
description of the data will be given. Relative accuracy among
the collected data strips will be examined. The absolute
accuracy procedure and results are also presented. The
procedure starts by selecting an appropriate area with some
specific characteristics, as will be discussed later, to conduct a
ground topographical survey as a reference for the assessment.
Collecting ground points was done using both GPS and typical
ground survey methods. A detailed analysis of the laser data
over that same area was performed. The results will include
both the planimetric as well as the height accuracy of the laser
data.
2. ERROR SOURCES IN LASER ALTIMETRY DATA
There are many sources of error and uncertainty that affect the
quality of the laser scanning data. They vary in their influence,
in the resulting error magnitudes, and in the way they should be
corrected or avoided. The resulting errors are an outcome of the
laser ranging computation, the scanning system, topography,
the atmosphere, positioning and navigation systems, and system
integration factors. Some of the major sources will be discussed
in brief with the way they should be treated to eliminate or at
least minimize their effects.
144
2.1 Laser unit and scanning system
The misalignment between the sent and the received pulse is
one of the error sources in range computation and point
positioning. Also the error in coincidence between the platform
coordinate system origin and the mirror center is another type
of misalignment error. This kind of error is correctable through
calibration (Morin and El-sheimy, 2002). Return signal
detection, range bin quantization, and the inaccuracies in the
pulse travel time measurement are another source of error in
range computation. These errors cannot be eliminated totally
but can be minimized by increasing the time resolution and
improving the synchronization between the clock and the laser
system.
2.2 Topography and atmosphere
Many examples can be cited under this category, only the
major ones will be mentioned. Vegetation and other objects
occluding the terrain introduce systematic error since they do
not represent the real terrain surface. Rough terrain and steep
slopes can generate artifacts in height measurements especially
with large footprints since the elevation error depends on the
slope angle and the planimetric position. This error will be
more severe when the flight direction is parallel to the slope
contour. On the other hand, error in height due to slope is less
affected if the flight direction is in the direction of the gradient
(Schenk, 1999). Another type of error in terrain can be
generated when the laser pulse hits the side of a vertical object,
which yields a misleading profile. The return signal amplitude
also plays a role in data accuracy. Some gaps in the data might
be produced if the returned pulse cannot be detected due to its
weakness. Specular reflections may also produce regions of
missing data. Uncompensated atmospheric conditions (air
pressure, temperature, and humidity) may also influence the
accuracy. Those errors are not correctable in a rigorous sense
but some of their values can be estimated by interpolation and
others can be minimized through careful mission planning and
operation.
Internatic
23 GP:
Unlike p
rely on t
depends
accurate
carrier pl
of gener:
centimete
positiona
GPS po
PDOP, g
can be a
least one
to be we
the other
position
orbits m
observing
Multipatl
more tha
observati
centimet
issue an
Kleusber
the numb
another s
(troposph
atmosphe
accuracy.
they can
Mikhail ¢
Error in
misalignr
the accel
and grav
significar
The data
ALTM 1.
on April |
with an
approxim
approxim
data con
direction.
and widtl
about 60
approxim
4.
Airborne
with a str
and the f]
and over!
covered. (
a mean
imprecisic
cause the
two diffe
