inbul 2004
ee
re shows
id that re-
[he over-
mera but
ges is in-
> scheme
with sub-
tzgibbon,
tion from
natic Ex-
32,
ards fully
altsavias,
kshop on
m Aerial
d dispar-
ints. In:
Rhodes,
Auto-
)amiyan,
s, IAPR,
w geom-
ess.
ellierung
opischen
ver.
. 3D Re-
s. PhD
juisition:
> Under-
AIRBORNE MONITORING OF VEHICLE ACTIVITY IN URBAN AREAS
U. Stilla? *, E. Michaelsen?, U. Soergel 5 S. Hinz?, J. Ender?
? Photogrammetry and Remote Sensing, Technische Universitaet Muenchen, 80290 Muenchen, Germany
br. . A . «fo .
FGAN-FOM Research Institute for Optronics and Pattern Recognition, 76275 Ettlingen, Germany
¢ Remote Sensing Technology, Technische Universitaet Muenchen, 80290 Muenchen, Germany
‘FGAN-FHR Research Institute for High Frequency Pysics and Radar Techniques, 53343 Wachtberg, Germany
stilla@bv.tum.de
Commission III, WG III/4
KEY WORDS: Urban, Monitoring, Aerial, Infrared, SAR
ABSTRACT:
In this paper several possibilities of vehicle extraction from different airborne sensor systems are described. Three major frequency
domains of remote sensing are considered, namely (i) visual, (ii) thermal IR and (iii) radar. Due to the complementing acquired
scene properties, the image processing methods have to be tailored for the peculiarities of the different kinds of sensor data.
(i)Vehicle detection in aerial images relies upon local and global features. For modelling a vehicle on the local level, a 3D-
wireframe representation is used describing prominent geometric and radiometric features of cars including their shadow region. A
vehicle is extracted by a “top-down” matching of the model to the image. On the global level, vehicle queues are modelled by
ribbons that exhibit typical symmetries and spacing of vehicles over a larger distance. Fusing local and global extractions makes the
result more complete. (ii) Particularly at night video sequences from an infrared camera yield suitable data to assess the activity of
vehicles. At the resolution of approximately one meter vehicles appear as elongated spots. However, in urban areas many additional
other objects have the same property. Vehicles may be discriminated from these objects either by their movement or by their
temperature and their appearance in groups. Using map knowledge as context, a grouping of vehicles into rows along road margins
is performed. (iii) The active scene illumination and large signal wavelength of SAR allows remote sensing on a day-night basis and
under bad weather conditions. High-resolution SAR systems open the possibility to detect objects like vehicles and to determine the
velocity of moving objects. Along-track interferometry allows estimation even small vehicle movements. However, in urban areas
SAR specific illumination phenomena like foreshortening, layover, shadow, and multipath-propagation burden the interpretation.
Particularly the visibility of the vehicles in inner city areas is in question. A high resolution LIDAR DEM is incorporated to
determine the visibility of the roads by a SAR measurement from a given sensor trajectory and sensor orientation. Shadow and
layover areas are detected by incoherent sampling of the DEM. In order to determine the optimal flight path a large number of
simulations are carried out with varying viewing and aspect angles.
1. INTRODUCTION
Traffic monitoring in dense build-up areas is a complex issue.
Research on this topic is motivated from different fields of
application: Traffic-related data play an important role in urban
and spatial planning, e.g., for road net optimization and for
estimation or simulation of air and noise pollution.
Furthermore, because of the growing amount of traffic, research
on car detection is also motivated by the strong need to
automate the traffic flow management by intelligent control and
guidance systems. The management of big events or the
emergencies after a disaster may require up-to-date traffic
information. This is important, e.g., to assess the parking
situation or to find ways for a task force into a danger zone
when crowds try to escape from this zone. Other fields of
application are found in the context of military reconnaissance
and extraction of geographical data for Geo Information
Systems (GIS), e.g., for site model generation and up-date.
The White Paper on the Common Transport Policy (2001) of
the European Communities has identified congestion, pollution
and energy consumption as key causes for the deteriorating
performance of Europe's transport systems, especially in the
industrialised urban regions. The 6th European Frame Program
* Corresponding author (stilla@bv.tum.de).
providing a budget of EUR 610 million for “Sustainable
Surface Transport” emphasizes the importance of traffic
monitoring and the demand on research. Some objectives are
environmentally friendly transport systems, safety improvement
and traffic congestion avoidance.
Automatic surveillance, planning, or control of traffic requires
data of the actual traffic situation. Stationary inductive or
optical sensors provide a permanent but only local measurement
of the traffic situation. Space borne sensors allow a spatial
measurement, but usually with low repetition rate and degree of
flexibility. Airborne systems can be used flexibly for data
recording of areas or routes. Depending on the task (e.g. traffic
census, measurement of traffic density or velocity, vehicle
discrimination, vehicle activity) different passive or active
sensors could be appropriate for the monitoring. For example,
investigations on congestion for a special route section have
been done with a video camera mounted on a helicopter [Gorte
et al, 2002] or vehicle discrimination in LIDAR data has been
studied [Toth et al., 2003]
In this paper, several possibilities to extract vehicles from data
of three frequency domains are described, namely (i) visual, (ii)
thermal IR and (iii) radar. Corresponding to the data, different
processing methods are demanded.