In: Stilla U, Rottensteiner F, Paparoditis N (Eds) CMRT09. IAPRS, Voi. XXXVIII, Part 3/W4 — Paris, France, 3-4 September, 2009
UTILIZATION OF 3D CITY MODELS AND AIRBORNE LASER SCANNING
FOR TERRAIN-BASED NAVIGATION OF HELICOPTERS AND UAVs
M. Hebel a , M. Arens a , U. Stilla b
a FGAN-FOM, Research Institute for Optronics and Pattern Recognition, 76275 Ettlingen, Gennany -
hebel@fom.fgan.de
b Photogrammetry and Remote Sensing, Technische Universität München, 80290 München, Gennany -
stilla@bv.tum.de
KEY WORDS: Airborne laser scanning, LiDAR, GPS/INS, on-line processing, navigation, city models, urban data
ABSTRACT:
Airborne laser scanning (ALS) of urban regions is commonly used as a basis for 3D city modeling. In this process, data acquisition
relies highly on the quality of GPS/INS positioning techniques. Typically, the use of differential GPS and high-precision GPS/INS
postprocessing methods are essential to achieve the required accuracy that leads to a consistent database. Contrary to that approach,
we aim at using an existing georeferenced city model to correct errors of the assumed sensor position, which is measured under non
differential GPS and/or INS drift conditions. Our approach accounts for guidance of helicopters or UAVs over known urban terrain
even at night and during frequent loss of GPS signals. We discuss several possible sources of errors in airborne laser scanner systems
and their influence on the measured data. A workflow of real-time capable methods for the segmentation of planar surfaces within
ALS data is described. Matching planar objects, identified in both the on-line segmentation results and the existing city model, are
used to correct absolute errors of the sensor position.
1. INTRODUCTION
1.1 Problem description
Airborne laser scanning usually combines a LiDAR device
(light detection and ranging) with high-precision navigational
sensors (INS and differential GPS) mounted on an aircraft.
Range values are derived from measuring the time-of-flight of
single laser pulses, and scanning is performed by one or more
deflection mirrors in combination with the forward moving
aircraft. The navigational sensors are used to obtain the 3D
point associated with each range measurement, resulting in a
georeferenced point cloud of the terrain. A good overview and a
thorough description of ALS principles can be found in (Wehr
& Lohr, 1999). Laser scanning delivers direct 3D measurements
independently from natural lighting conditions, and it offers
high accuracy and point density.
A well-established application of laser point clouds acquired at
urban areas is the generation of 3D city models. However, the
overall precision of the derived city model highly depends on
the accuracy of the data input, which is directly dependent on
the exactitude of the navigational information. Great efforts are
usually required during data acquisition and postprocessing in
order to achieve high fitting accuracy of multiple ALS datasets
(e.g. neighboring strips). While ALS data acquisition is
commonly done to supply other fields of studies with the
necessary data, few examples can be found where laser scanners
are used directly for pilot assistance. One of these examples is
the HELLAS obstacle warning system for helicopters (Schulz et
al., 2002), which is designed to detect wires and other obstacles
for increased safety during helicopter missions.
Despite increasing performance of LiDAR systems, most remote
sensing tasks that require on-line data processing are still
accomplished by the use of conventional CCD or infrared
cameras. Typical examples are airborne monitoring and
observation devices that are used for automatic object
recognition, situation analysis or real-time change detection.
Utilization of these sensors can support law enforcement,
firefighting, disaster management, and medical or other
emergency services. At the same time, it is often desirable to
assist pilots with obstacle avoidance and aircraft guidance in
case of poor visibility conditions, during landing operations, or
in the event of GPS dropouts. Three-dimensional information as
provided by the LiDAR sensor technology can ease these tasks,
but the existence of differential GPS ground stations and the
feasibility of comprehensive data analysis are not to be
considered for these real-time operations.
1.2 Overview
The approach of using ALS information to provide on-line
navigation support for aircraft guidance over urban terrain is
opposite to the process of city model generation. In contrast to
the demand for high-precision positioning techniques, it is
assumed that a proper georeferenced city model is already
available. This database can be used to generate a synthetic
vision of the terrain according to current position and
orientation of the aircraft. Moreover, ALS measurements and
matching counterparts in the city model can be taken into
consideration if additional navigational information is needed,
for example in cases of degraded GPS positioning accuracy.
This paper presents a workflow of methods for the segmentation
of planar surfaces in ALS data that can be accomplished in line
with the data acquisition process. Since most of currently used
airborne laser scanners, like the RIEGL LMS-Q560, measure
range values in a pattern of parallel scan lines, the analysis of
geometric features is performed directly on this scan line data.
Straight line segments are first segmented and then connected
across consecutive scan lines to result in planar surfaces. All
proposed operations are applicable for on-line data processing.
