Skip to Main Content
Automatic traffic monitoring and surveillance have become essential for effective road usage and management. Various sensors have been used to estimate traffic parameters, but their installation and maintenance is often difficult and costly. Among the technologies being investigated, computer vision promises the most flexible and reliable solutions to estimate traffic parameters. This paper proposes a wireless sensor network (WSN) architecture for autonomous traffic monitoring, based on computer vision techniques for automatic scene analysis and interpretation. The paper first discusses the motivation for the work and the relevant design issues. Then, the proposed architecture and the relevant modules are described in detail. Finally, experimental results are shown, which prove the accuracy of the proposed approach.