Automatic number-plate recognition: neural network approach

Abstract

Automated number-plate recognition, using image processing techniques, is indicating new ways of capturing traffic data. Recent advances in computer vision technology and the falling prices of the related devices encourage the use of video/ closed-circuit television cameras by making it practical to automatically identify vehicles visually on-line or off-line. Automatic vehicle identification system could be achieved using automated number-plate reading with high accuracy and low processing time. The objective of this research is to explore the potential of using the BAM neural network for number-plate reading. Image processing techniques are used to extract the location of each included character, and the BAM neural network is applied to the character recognition process.