Video-Based Automated Identification of Freeway Shoulder Events

Abstract

Transportation agencies have increasingly turned to information technology to help address the growing challenges of managing congested urban freeways. Although off-the-shelf information technology products, such as closed-circuit television (CCTV) systems and video image vehicle-detection systems (VIVDS), have proved very useful, their full potential cannot be achieved without integrating the products. The purpose of this research effort was to build on an existing prototype of an integrated CCTV–VIVDS system to provide automatic shoulder event identification functionality. This paper presents the design of the integrated system, provides details on the tools and effort required to support integration of the information technology elements, and describes the results of field testing of the system.