Abstract

This article presents a license plate number recognition system vehicle for Turkish license plates. This article proposed steps for an improved Bernsen algorithm applied for the plate number recognition system. The proposed system is designed to extract the characters features of vehicle license plate numbers automatically from digital images. Also, this proposed technique is of three stages, the first is the detection technique which is based on an improved Bernsen algorithm, while the second technique is the segmentation of character in which Vertical and horizontal projection analysis techniques are used. The last stage is to recognize the characters, in which the letters and numbers are classified using an artificial neural network (ANN). In this article, a sample of 100 Turkish vehicle license plates were captured and tested, from the experimental result shows how the algorithm segment and recognize the characters very effectively.

Keywords

License Plate Recognition (LPR), Optical Character Recognition, Neural Network, Image Processing, Machine Learning.