一种基于多重特征的车牌定位算法

车牌自动识别是现代智能交通的重要组成部分,而车牌定位技术又是车牌自动识别系统的核心之一,然而对于真实的复杂背景下的图像,现存的基于特征的定位算法依然存在适应性差、鲁棒性不强的问题。提出了一种基于多重车牌特征的行列扫描线的定位算法,突破了原有的基于特征的行,列扫描算法中容易出现车牌区域错误判断的瓶颈。对于真实的具有复杂背景的图像进行实验后,得到了理想的定位效果。

Approach of License Plate Extraction Based on Multi-characteristic

License Plate Recognition(LPR)plays an important role in Intelligent Transportation System and the location of the license plate is the critical technique in the LPR.However,the current method of license plate location,which is based on the characteristics of the plate,is still not adaptive and robust when used for the images having complicated background.The paper introduces a new license plate extraction method which is based on the multi-characteristic of the plate.This method breaks the bottleneck that the old characteristic-based license plate extraction methods with row and column scanning always neglect and mistake the plate area.The experiment shows that it gets ideal effect on the great amount of images having complicated background.