Abstract
From the background information supplied by the inventors, news correspondents o btained the followingquote: “Identification of vehicles using computer-implemen ted image analysis is used in various applications. For example, license plate I mage recognition systems are frequently used in automatedtollgates that identif y moving vehicles to automatically collect tolls. In addition, systems that comb inelicense plate recognition and detection of vehicle features, such as car col or or make, are used to improveidentification accuracy when the license plate c annot be adequately identified. Indeed, highly accuratevehicle identification b ased on only vehicle features is possible with techniques based on machine-learning algorithms such as convolutional neural networks. These algorithms capture f eatures from images ofknown vehicles and then identify an unknown vehicle in an image by correlating image features. Althoughthese techniques may be computer- intensive, they can improve identification accuracy and facilitate theautomatio n of a vehicle image identification system.
NETL
NSTL