An Integrated Number Plate Recognition System through images using Threshold-based methods and KNN
V. Uma Maheswari, Rajanikanth Aluvalu, Swapna Mudrakola
Abstract
In the last few decades, the use of vehicles in our daily life has become mandatory and increased drastically. Sometimes, controlling traffic and identifying vehicle owners manually becomes tedious due to crowd signals, which disobey the traffic rules and drive fast and abnormal. This demands an efficient and automatic system to solve the problem these days. Still, it is challenging in such cases as moving vehicles fast, font on number plate, illumination, etc. This led to developing efficient and automatic number plate detection as the solution. This paper presents automatic number plate detection with number diagnosis and tracking by applying various methods such as thresholding, morphological methods, contour detection, etc. Later, KNN is used for classification to improve accuracy. The proposed method tested on datasets DB1 and DB2 proves better in terms of accuracy, recognition rate, and retrieval rate.