基于HSI空间和边缘检测的一种高性能的车牌定位方法  

High Performance License Plate Location Method Based on HIS Space and Edge Detection

作　　者：陈财明[1] 

机构地区：[1]台州学院物理与电子工程学院,临海317000

出　　处：《科学技术与工程》2009年第15期4372-4376,共5页Science Technology and Engineering

摘　　要：首先把图像的RGB空间转换成HSI空间,根据车牌区域在HSI空间的特征并结合模糊集合理论可分别得到H特征图、S特征图和I特征图。如果根据这三幅特征图定位车牌失败,则根据车牌的彩色边缘得到边缘特征图。如果边缘特征图也无法定位车牌,则把HSI特征图和边缘特征图结合起来再对车牌重新定位。进行的实验中共有200辆车辆图像,只有9辆定位失败,准确率达到95%。The RGB color model of the car image will be converted into the HSI model from which H characteristic map, S characteristic map and I characteristic map can be gotten according to the characteristics of the license plate in the HSI model. If the license plate can not be located by using these three maps, then the color edge characteristic map can be gotten from the RGB model. If the license plate cannot be located either, then the HSI fuzzy map and edge characteristic map should be integrated to find the license plate. There are 200 vehicles for the experiment. Only 9 vehicles failed to find the plates. The accuracy is 95 %.

关 键 词：车牌定位 HSI空间 边缘检测 模糊集合 

分 类 号：TF391.41[冶金工程—冶金机械及自动化]