Publication: Automatic vehicle counting from video for traffic flow analysis
Program
KU Authors
Co-Authors
Publication Date
Language
Embargo Status
Journal Title
Journal ISSN
Volume Title
Alternative Title
Abstract
We propose a new video analysis method for counting vehicles, where we use an adaptive bounding box size to detect and track vehicles according to their estimated distance from the camera given the scene-camera geometry. We employ adaptive background subtraction and Kalman filtering for road/vehicle detection and tracking, respectively. Effectiveness of the proposed method for vehicle counting is demonstrated on several video recordings taken at different time periods in a day at one location in the city of Istanbul.
Source
Publisher
IEEE
Subject
Computer science, artificial intelligence, Computer science, theory and methods, Engineering, electrical and electronic, Engineering, mechanical
Citation
Has Part
Source
2007 IEEE Intelligent Vehicles Symposium, Vols 1-3