Academic Editor: Youssef EL FOUTAYENI
Received |
Accepted |
Published |
30 January 2020 |
14 February 2020 |
10 March 2020 |
Abstract: Road and highway traffic remains a subject of study for the public authorities because it is at the heart of many concerns such as mobility and security. We propose in this work a system for detecting, tracking and estimating vehicle trajectories and continuous line crossing for possible accident prevention. We used kalman filter known by its performance in motion detection. The main phases adopted consist at first in calculating of the background image that will be used to detect the moving object by subtracting from the current frame. The detected blobs will eventually have a processing set such as merging adjacent regions and elimination of negligible regions. The characteristics of the blobs detected will then be processed by the kalman filter, to estimate the trajectory of vehicle in the following frames. The tests carried out on several roads and highway scene have given very satisfactory results on detecting and tracking vehicle and detecting of continuous line crossing that can be exploited for accident prevention