A Smart Camera for Traffic Surveillance
Abstract — The integration of advanced CMOS image sensors with high-performance
processors into an embedded system facilitates new application classes such as smart
cameras. A smart camera combines video sensing, video processing and communication
within a single device. This paper reports on the prototype implementation of
a smart camera for traffic surveillance. It captures a video stream, computes traffic
information and transfers the compressed video stream and the traffic information
to a network node. The achieved experimental results of the implemented stationary
vehicle detection demonstrate the feasibility of our approach.
1 Introduction
Due to their logarithmic behavior, high dynamic range and high bit resolution the low-cost
and low-powerCMOS sensors acquire images with the necessary quality for further image
processing under varying illumination conditions. The integration of these advanced image
sensors with high-performance processors into an embedded system facilitates new
application classes such as smart cameras. Smart cameras not only capture images or
video sequences, they further perform high-level image processing such as motion analysis
and face recognition on-board and transmit the (compressed) video data as well as the
extracted video information via a network.
An important application area where smart cameras can potentially and advantageously
replace most known cameras, frame grabbers and computer solutions is visual traffic
surveillance [1]. CMOS image sensors can overcome problems like large intensity contrasts
due to weather conditions or road lights and further blooming, which is an inherent
weakness of existing CCD image sensors. Furthermore, noise in the video data is reduced
BRAMBERGER, PFLUGFELDER, MAIER, RINNER, STROBL, SCHWABACH
by the capability of video computation “close” to the CMOS sensor. Thus, the smart
camera delivers a new video quality and better video analysis results, if it is compared
to existing solutions. Beside these qualitative arguments and from a system architecture
point of view, the smart camera is an important concept in future digital and heterogeneous
third generation visual surveillance systems [2]. Not only image enhancement and image
compression but also video computing algorithms for scene analysis and behavior understanding
are becoming increasingly important. These algorithms have a high demand for
real-time performance and memory. Fortunately, smart cameras can support these demand
as low-power, low-cost embedded systems with sufficient computing performance
and memory capacity. Furthermore, they offer flexible video transmission and computing
in scalable networks with thousands of cameras through a fully digital interface.
The purpose of this paper is to present first results of an ongoing research project between
ARC Seibersdorf research, the Institute for Technical Informatics at Graz University
of Technology and the pattern recognition and image processing group at Vienna
University of Technology. The primary goal of this project is the development of a smart
camera for traffic surveillance. This paper presents the camera’s prototype implementation
and a case study of our smart camera concept with respect to stationary vehicle
detection in tunnels. We chose this application, because it is by far the most important
application (80 percent of all applications) in traffic surveillance.
The remainder of the paper is organized as follows: Section 2 presents the requirements
of a smart camera and lists the related work. The hardware architecture is outlined in
Section 3, while the software components required by the smart camera are depicted in
Section 4. Experiments and a prototype description is described in Section 5. Section 6
concludes this paper with a short discussion.
2 Requirements and Related Work
2.1 Requirements of a Smart Camera
In general a smart camera is compromised of a sensor, a processing and a communication
unit. In this section we briefly discuss the requirements for each of these units as well as
some system wide requirements.
2.1.1 Sensor Requirements
The image sensor is the prime input for a smart camera. An appropriate image quality is,
therefore, essential for the performance of the entire system.
Dynamic Range Traffic surveillance applications enforce high demands on the image
sensor. Typical traffic situations may contain a high dynamics, e.g., when high-intensity
areas, such as the high-beam of a vehicle, appear concurrently with low-intensity areas
such as the car’s silhouette at night. Image sensors with high dynamic range and little
blur are preferred for these applications. Additionally, high dynamic-range sensors ease
the design of the camera control and the control of the lens aperture in changing light
conditions.
download full report
http://vmars.tuwien.ac.at/wises/papers/1...ection.pdf