EVALUATION OF INDUCTIVE-LOOP EMULATION
ALGORITHMS
FOR UTC SYSTEMS
Luigi Di Stefano AND Enrico Viarani: DEIS, University of Bologna
Viale Risorgimento 2, 40136 Bologna, Italy e-mail
Keywords: Computer Vision, Traffic Monitoring, Inductive Loop, Cross
Correlation
Abstract
The paper presents the recent results of a research project active
at the University of Bologna and aimed at the emulation by means of
computer
vision algorithms of the inductive sensors used for vehicle detection
in
Urban Traffic Control (UTC) systems. Emulation of inductive sensors
using
computer vision has traditionally been based on Sum of Absolute
Differences
(SAD) algorithms. In this paper we propose an approach based on the
Normalized
Cross Correlation function (NCC) and present experimental results aimed
at comparing SAD and NCC based vehicle detection. Our results show that
the NCC-based algorithm significantly outperforms the SAD-based one.
Get the paper file