小生境遗传算法在多车牌定位中的应用  

License plate location based on niche genetic algorithm

作　　者：阴亚芳[1] 张婧琪[1] 廖延娜[2] 

机构地区：[1]西安邮电大学电子工程学院,陕西西安710121 [2]西安邮电大学理学院,陕西西安710121

出　　处：《西安邮电大学学报》2015年第1期40-43,共4页Journal of Xi’an University of Posts and Telecommunications

基　　金：陕西省教育厅科学研究计划资助项目(12JK0559)

摘　　要：为了在复杂背景下对多车牌进行准确定位,提出一种基于小生境遗传算法的多车牌定位方法。利用一组一维滤波器对图像平滑处理得到图像的特征向量,建立小生境遗传算法模型,设计适应度函数,并确定小生境算法机制,实现多车牌定位。仿真实验表明,基于小生境遗传算法的多车牌定位方法能够准确地对多车牌进行定位,平均定位率达到90%以上。In order to locate multiple license plates in complex background accurately,a multiple license plate location method based on niche genetic algorithm is proposed. Firstly, the feature vector of the image is obtained by a different set of uni-dimensional filter. Secondly, a niche genetic algorithm model is established, a fitness function is designed, and a mechanism of niche algorithm is determined. Finally, the license plate location is achieved. Simulation results show that this license plate location method based on niche genetic algorithm can locate the license plate accurately. The average localization rate is over 90%.

关 键 词：车牌定位 小生境 遗传算法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]