Robust Vehicle and Traffic Information Extraction for Highway Surveillance
- PDF / 1,361,171 Bytes
- 17 Pages / 600 x 792 pts Page_size
- 34 Downloads / 198 Views
Robust Vehicle and Traffic Information Extraction for Highway Surveillance Akio Yoneyama Multimedia Communications Lab., KDDI R&D Laboratories Inc., 2-1-15 Ohara, Saitama 356-8502, Japan Email: [email protected]
Chia-Hung Yeh Department of Electrical Engineering and Integrated Media Systems Center, USC Viterbi School of Engineering, University of Southern California, Los Angeles, CA 90089-2564, USA Email: [email protected]
C.-C. Jay Kuo Department of Electrical Engineering and Integrated Media Systems Center, USC Viterbi School of Engineering, University of Southern California, Los Angeles, CA 90089-2564, USA Email: [email protected] Received 1 January 2004; Revised 13 January 2005 A robust vision-based traffic monitoring system for vehicle and traffic information extraction is developed in this research. It is challenging to maintain detection robustness at all time for a highway surveillance system. There are three major problems in detecting and tracking a vehicle: (1) the moving cast shadow effect, (2) the occlusion effect, and (3) nighttime detection. For moving cast shadow elimination, a 2D joint vehicle-shadow model is employed. For occlusion detection, a multiple-camera system is used to detect occlusion so as to extract the exact location of each vehicle. For vehicle nighttime detection, a rear-view monitoring technique is proposed to maintain tracking and detection accuracy. Furthermore, we propose a method to improve the accuracy of background extraction, which usually serves as the first step in any vehicle detection processing. Experimental results are given to demonstrate that the proposed techniques are effective and efficient for vision-based highway surveillance. Keywords and phrases: traffic monitoring, object tracking, moving cast shadow, occlusion, nighttime detection, background subtraction.
1.
INTRODUCTION
Vision-based traffic monitoring systems are widely used in intelligent transportation systems (ITS). The goal of a traffic monitoring system is to extract traffic information, such as the vehicle count, traffic events, and traffic flow, which plays an important role for traffic analysis and traffic management. Several different types of devices, including loop detectors, sensors, and cameras, have been employed in traffic monitoring systems. Recently, vision-based analysis systems have become popular in transportation management due to their capability to extract a wide variety of information in comparison with the sensor-based system. Vision-based systems have a good potential for highway surveillance applications [1, 2], and useful traffic information such as vehicle dimensions, lane changes, and other traffic-related information can be effectively extracted. However, it is challenging to maintain detection accuracy at all time since vision-based processing is sensitive to environmental factors such as lighting, shadow, and weather
conditions. The following factors tend to result in the degradation of detection performance. (i) Shadow Moving region extraction is one of the fundamental steps in object detect
Data Loading...
