Printer-friendly version

Traffic Surveillance > Road-based Technologies > Roadside >

Video-Image Detection

• Video image detection systems (VIDS) employ machine vision technology to automatically analyze traffic data collected with Closed Circuit Television (CCTV) systems.
• VIDS can be used to monitor freeway conditions, arterials and intersections, detect incidents and classify vehicles.
• A distinction should be made between VIDS that can actuate signals and detect incidents, queue lengths and other parameters with little or no human intervention and video surveillance that is mostly used for "manual" inspection of existing traffic conditions and verification of incidents. Most existing video feeds are used in a manual fashion. However, VIDS are rapidly and successfully being deployed to perform a variety of functions. See our Telecommunications Diagram of CCTV for more information.

Key Results

Most video networks used by the ITS community, and other industries, are analog networks typically based on Frequency Division Multiplexing (FDM) transmission techniques. Analog video systems are a mature, proven technology that yields high quality video at relatively low cost. The main advantage of VIDS lies in its wide area detection capabilities, usually several lanes for one camera. This allows the efficient detection of shock waves and other spatial traffic parameters, such as density, queue lengths, and speed profiles, that cannot be easily obtained by conventional devices. In addition, VIDS can be used in a number of different ways outlined below. Benefits

• Rapid incident detection.
• Identification of the incident type, the level of gravity and what type of intervention is needed if any. This is mostly done by human operators.
• Vehicle classification, intersection monitoring, signal actuation and license plate reading (which can be used for enforcement and travel time estimation).
• Collection and potential analysis of traffic and incident data. Information provided in the video sequences immediately preceding an incident can be particularly valuable for understanding where and how incidents occur so as to enhance infrastructure reliability. Surveillance data could be used not only for safety but also for transportation planning, operations and research.
• Enhanced installation safety: because installation typically does not require lane closures, traffic personnel safety is enhanced and traffic disruptions are minimized.
• Integration with other existing surveillance technologies as well as with other management strategies such as variable message signs and adaptive ramp metering control.
• Traveler information in the form of video feeds accessible on numerous web sites allows travelers to make better-informed travel decisions and to avoid non-recurrent congestion as well as inclement weather.

• Initial capital outlays may exceed the costs of traditional detection methods such as loop detectors but VIDS have lower installation and life-time (operations and maintenance) costs than traditional detection methods. Capital outlays include communications costs, which can be significant for digital video technology. Digital video technology is close to three times more expensive than traditional analog technology.
• With an existing transmission link, a color CCTV camera with a 20-year lifetime can vary in cost between $10,000 and $50,000 with annual operations and maintenance costs of between $200 and $1,000. A CCTV camera tower with a 20-year lifetime, can vary in cost between $18,000 and $50,000 with maximum annual operations and maintenance costs of $900. This does not include the software systems and algorithms required for automatic surveillance .

• In the last few years complex automatic incident detection (AID) and queue estimation systems have been implemented either as tests or as actual deployments. These systems automatically detect incidents, estimate queues and change variable message signs to warn travelers of problems.
• Given its high quality-to-price ratio, analog video transmission technology will continue to be deployed for some years. However, despite the efficiency of analog systems, digital video systems are gaining in importance and are widely believed to become the dominant technology for video transmission in ITS over the next 10 to 15 years. Eventually, digital networks and systems will supersede analog systems everywhere, so it seems to make sense to invest in digital video equipment. However, the transition to digital video transmission equipment will only happen gradually because of costs and the need for an industry-wide video compression standard (MPEG-2 seems to be the emerging standard).

• There are no technological barriers given the technical maturity of video systems. The main technical challenge lies in refining automatic incident detection algorithms and software systems.
• Regarding digital video, the chief barriers to implementation are institutional (in particular the proper planning of communications infrastructure) and monetary (the high cost of setting up digital networks).

• There are over 5,000 VIDS in existence today throughout the world.
• Simple VIDS, for signal actuation, speed estimation or traffic counts, are implemented throughout Europe, the United States and Australia. These are present on a more limited basis in Asia (principally in China and Japan).
• More complex VIDS that automatically detect incidents, estimate queue lengths and actuate variable message signs are less widespread. Examples include:
  • The Migrazur system on the Escota Network of 430km of freeways in the southern Provence-Alpes-Cote d'Azur regions of France.
  • DIVA system in the Les Halles tunnel in central Paris.
  • The A4-A31/Brescia-Padova freeway in Northern Italy.
  • The City Link project in Sydney, one of the biggest and most sophisticated projects covering 22km of high-quality roads.

   

Author: Dimitri Loukakos