Vehicle Plate Matching using License Plate Recognition based on Modified Levenshtein Edit Distance

Vehicle License plate recognition (LPR) method is a fullgrown so far deficient approach used for computerized toll group and rapidity enforcement. Recently, an sophisticated matching approach that combines Bayesian likelihood and Levenshtein text-mining method was planned to improve the exactness of computerized vehicle license plate matching. The key module of this technique is what we known as PrecisionRecall curve, which contains the conditional probabilities of observing one character at one node for a given observed character at an additional station. Therefore, the evaluation of the performance constraint relies on the by hand extracted position truth of a large number of plates, which is an unwieldy and deadly process. To beat this negative aspect, in this cram, we propose an inventive novel LPM-MLED (License Plate Matching Modified Levenshtein Edit Distance) method that removes the need for extracting ground truth by hand. The propose method perform well in the exactness in plate matching, in comparison with those generated from the meticulous manual method. Furthermore, this method outperforms their manual counterparts in plummeting false matching rates. The computational LPMMLED technique is also cheaper and easier to implement and continues to improve and correct itself over time.