Detecting License Plate based on Wavelet Transform from Complex Scene

D. Liu and M. Xie (PRC)

Keywords

Image processing, License plate location, Gradientvariance, Wavelet transform

Abstract

Detection of vehicle license plate is the most fundamental process in Vehicle License Plate Recognition System. This paper presents a license plate detection algorithm from complex background based on wavelet transform. The proposed algorithm consists of two main modules: license plate regions roughly detecting and license plate accurately locating. The former characterized by vertical gradients detection extracts candidate regions from an input image, while the latter aims to locate the license plate fast and accurately in terms of wavelet transform and high frequency feature of license plate. Experiments have been conducted on 353 images taken from various scenes and different conditions, including diverse angles, different lightening conditions and the dynamic conditions. And the algorithm can quickly and correctly detect the region of vehicle license plates. The rate of success is 98.02% with 7 images failed.

Important Links:



Go Back