LICENSE PLATE RECOGNITION

License Plate Recognition (LPR) technology has been used for off-line automobile enforcement purposes. The technology has seen mixed success with correct reading rate around 60 to 70% depending on the specific application and environment. This limitation can be, and is often, remedied through human verification after the fact and before a citation is issued. Armed with advanced text-mining algorithms, this study enables LPR technology for real-time enforcement by matching plates whether correctly or incorrectly read at various locations in a network or along a corridor and, hence, tracking the movement and speed of vehicles. The focus of the project is on heavy vehicles as they are required to enter weigh stations, where the LPR tracking information can be used, in real time, for speed enforcement and/or as a triggering factor for other inspectional activities. The first phase of the project, which is reported herein, saw the development of the enabling text-mining algorithm and the demonstration of plate matching rate improvement from less than 60% to better than 90%. The second phase will deploy the LPR technology and matching algorithm in real-time so that a trooper sitting in the weigh house can warn a truck driver, in person and within minutes, of his or her speeding offense without having to chase the truck down in a patrol car. The initial objective of the study is to devise an inexpensive and effective means for helping improve air quality in non-attainment metropolitan areas through speed enforcement. The successful deployment of such a measure can also potentially lead to improved highway safety, fuel efficiency, and national security.