Robust and adaptive traffic surveillance system for urban intersections on embedded platform

Abstract

The paper presents a real time traffic monitoring system that makes use of image processing algorithm to detect and estimate the of count of vehicles using motion detection approach. Traffic congestion is a serious issue, which is the root cause of a series of serious problems. Conventional traffic light controllers have limitations because they make use of the predefined hardware, whose functioning is governed according to program that does not have the flexibility of modification on real time basis. The proposed system makes use of a differential algorithm in order to determine the signaling duration of each lane of intersection. The system provides different delays for different junctions thus optimizing the waiting time of each user. This flexibility of timing and controlling prevents the congestion of vehicles in squares due to high waiting time for the green light. The traffic flux density determines the effective number of vehicles at any intersection and hence is critical in allocation of signaling duration to any intersection. Real time analysis presents many challenges in video analysis and in order to lower down the computational complexities, the algorithm makes use of simple background subtraction technique. The system has been tested for a number of video sequences. The results produced are extremely encouraging and hence the system can be applied in real time traffic control in urban areas.

Traffic Signal Control System Using Contour Approximation Deep Q-Learning

“…One such instance is the Internet of Things [3,4] where traffic is efficiently controlled using cameras and other sensors. Wireless sensor networks (WSN) [5] and the use of embedded platforms [6] are two more effective techniques.…”