Automated Traffic Signal Prediction from Surveillance Videos

Karthick.S Karthick.S, Deeban.B Deeban.B, S. Abirami
2012 International Journal of Computer Applications  
The traffic signals available in the present are based on the static feed of time without considering the actual available traffic. This leads to a situation where vehicles wait unnecessarily in one of the lanes while the traffic flow is not up to the considerable amount in the other lane. This paper provides a system to monitor the traffic flow automatically in traffic signals where video cameras are fixed. The time feed is made dynamic and automatic by processing the live traffic videos. The
more » ... raffic videos. The time for the signal is determined by two main factors: based on the density of the vehicles and on the number of vehicles in the lane. These inputs are given to a proposed algorithm which determines the optimal time period for the signal. General Terms Image processing, Traffic signal automation. Keywords Dynamic traffic signal prediction, density of vehicles, vehicular count. Section 2 gives system architecture design and explains the activities which are taking place in the system. Section 3 presents the various test cases and their results. Section 4 analyses the performance of the system. Section 5 concludes the work and gives suggestions for the future work. Real 75 Reported 75 Accuracy
doi:10.5120/5659-7554 fatcat:ducqj2e2unbwvk5firlhjwatjy