Adaptive CCTV

One of the aims of the Highways Agency (HA) is to improve the reliability of journeys through the reduction or better management of congestion.  Trials have shown that commercial CCTV automatic incident detection (AID) products, when used in isolation, generate too many false alerts to be useful to operators.

Cambridge Consultants has been engaged by the HA to assist in the development of a system, utilizing the HA’s currently installed CCTV network, to improve the speed with which traffic incidents were detected and cleared thereby reducing incident related congestion.

Adaptive CCTV makes use of a number of data sources, both live and historical, and applies a data fusion algorithm to combine the data in a probabilistic fashion.  The system analyses this data including real time video feeds from existing CCTV cameras to select those video feeds with the greatest probability of displaying an incident. The video feeds from these cameras are then highlighted, to the operators on a video wall, allowing faster detection of incidents by the operators.

An implementation of Adaptive CCTV has been live in the HA’s West Midlands RCC since late 2008.  It is currently serving as a tool for eliciting feedback from the control room operators and for evaluating the performance of the system.

The system has allowed operators to detect incidents in view of the system on average 3 minutes faster, this improvement in response times reduces the impact on journey times caused by the incidents.

For further information about data fusion analysis click here