Overview Motivation State of the Art Implementation Challenges Results References

Current State of the Art


        Video and photo identification software for license plate recognition (LPR) already exists, however it is often very narrow in its use. The majority of the software has been built for specific applications. The most well known, and potentially most disliked, is to identify a car on a tollway and to charge the owner a fee for passing through. For Illinois tollways, the company responsible for the project is J. A. Watts, Inc. The setup of their project generates an image where the license plate is at roughly the same location, size, and position in each image. From that, they pull the data necessary to charge the owner of the car. In cases where the plate cannot be read, the image is flagged and sent to a human reviewer. They do not provide statistics on their success rate, however, as indicated in a report by the National Law Enforcement and Corrections Technology Center, true success rates for modern LPR technology sit between 80 and 85 percent.

        Many other companies currently exist making money solving the problem of license plate recognition. One company, with perhaps more robust image recognition technology, is gtechna. The company creates and sells products for a variety of uses. Their software is able to recognize plates when mounted on moving law enforcement vehicles, for parking operators, and in a variety of other settings. Other major companies involved in this business include Raytheon, PlateSmart Technologies, and Genetec. Other private companies exist strictly in software development, often selling their products for upwards of $1,000.

        The industry is full of potential competitors in this market, however, much of this technology still remains quite expensive and performs at varying success rates.