Highlights ► Improved text-mining algorithms to process LPR data for real-time vehicle tracking. ► Overcame the challenge of matching strings without known database or dictionary. ► Field-tested procedures on thousands of speeding trucks at multiple Interstate sites. ► Achieved high plate matching rate (95%) inspite of low LPR reading rate (60%). ► Reduced and kept false matching rate very low (1%) with the proposed methodology.

    Abstract License-plate recognition (LPR) technology has been widely applied in many different transportation applications such as enforcement, vehicle monitoring, and access control. Recently, there has been effort to exploit an LPR database for vehicle tracking using popular template matching procedures. Existing template matching procedures assume that the true reference string is always available. However, under a two-point LPR survey, a vehicle could have its plate misread at both locations generating a pair of misread strings (or templates) with no reference for matching. To compensate for LPR misreading problem, we propose a new weight function based on a probability model to match the observed outcomes of a dual LPR setup. Also, considering that reversal errors are never made in LPR machines, new editing constraints as a function of the string lengths are proposed to avoid compensation for reversal errors. These editing constraints are incorporated into the constraint edit distance formulation to improve the performance of the matching procedure. Finally, considering that previous template matching procedures do not take advantage of passage time information available in LPR databases, we present an online tracking procedure that considers the properties of probability distribution of vehicle journey times in order to increase the probability of correct matches. Experimental results show that our proposed procedure can improve the accuracy of LPR systems and achieve up to 97% of positive matches with no false matches. Further research is needed to extend the ideas proposed herein to plate-matching with multiple, i.e., more than two, LPR units.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Online license plate matching procedures using license-plate recognition machines and new weighted edit distance


    Contributors:


    Publication date :

    2011-11-14


    Size :

    15 pages




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    English






    License Plate Surveillance System Using Weighted Template Matching

    Kim, Y. / Ko, M. | British Library Conference Proceedings | 2004


    LICENSE PLATE RECOGNITION SYSTEM AND LICENSE PLATE RECOGNITION METHOD

    CHEN SHU-HENG / LIAO CHIH-LUN / SHEN CHENG-FENG et al. | European Patent Office | 2020

    Free access