An Efficient Method for License Plate Localization Using Multiple Statistical Features in a Multilayer Perceptron Neural Network

Accurate license plate localization is the most important prerequisite in ANPR (Automatic Number Plate Recognition) systems. Majority of the existing algorithms use a single feature to obtain the license plate location which causes to potential false detections. In this article we propose a robust methodology using 16 statistical features while we still preserve real-time processing of the system which is a requirement for such applications. The proposed method uses a Vertical Projection technique and Discrete Fourier Transform (DFT) in order to extract multiple statistical features, as well as K-means clustering and multilayer perceptron neural network technique to identify the location of a license plate in an image. The method is compared with the state-of-the-art research in the field and the effectiveness of the method is evaluated for various types of license plates with different scripts.