01370287

Component

Video cameras play an important role in achieving the potential promised by Intelligent Transportation Systems. This paper proposes an improved Camshift method for automatic multi-vehicle tracking using video cameras for dealing with the complex traffic scenes such as target non-uniqueness, illumination variation, and color distraction in real traffic situations. The proposed method firstly uses a background subtraction approach for automatically detecting, selecting, and initializing the targets. Moreover, a motion estimation approach is proposed for estimating the motion states of the vehicles and predicting their new locations. Finally, a mixed model is established for searching the matching vehicles in the next image frame. Compared to the traditional Camshift method, the proposed method is efficient for accurately tracking vehicle movements, and herein is promising for real applications.

This paper was sponsored by TRB committee ABJ35 Highway Traffic Monitoring

01362476

12-2268

English

18p

2012

Transportation Research Board 91st Annual Meeting

Digital/other

Figures; Photos; References; Tables

Algorithms; Automatic vehicle identification; Computer models; Intelligent transportation systems; Tracking systems; Traffic surveillance; Video cameras

Data and Information Technology; Highways; Operations and Traffic Management; I70: Traffic and Transport; I73: Traffic Control

Transportation Research Board Annual Meeting 2012 Paper #12-2268

TRIS, TRB

Feb 8 2012 5:08PM