基于蓝色车牌背景的车牌精确定位实现方法  

Implementation Method of License Plate Accurate Location Based on Blue Background Information

作　　者：李强[1] 汪洋[1] 李伟博[1] 孙瑞杰[1] 

机构地区：[1]重庆邮电大学信号与信息处理重庆市重点实验室,重庆400065

出　　处：《电视技术》2011年第15期4-6,33,共4页Video Engineering

基　　金：信号与信息处理重庆市重点实验室建设项目(CSTC;2009CA2003);重庆市自然科学基金项目(CSTC;2010BB2411)

摘　　要：先对抓拍的彩色车牌图像进行车牌的粗定位,分割出车牌区域。对车牌区域做图像增强和滤波处理后,转化成二值图像,去除车牌的上下和左右边框,实现车牌的精确定位。针对抓拍的车牌彩色图像因光照不足的原因造成垂直粗定位不准的问题,提出了改进算法,提高了精确定位的准确率。The license plate is rough located in color image photographed and the area of license plate is segmented. Then, the color license plate image is transformed to binary images after image enhancement and filtering. Finally, the license plate is precisely located after the upper and lower and left and right border of license plate are removed. An improved vertical rough location algorithm is proposed in view of insufficient light color license plate image and increase the accurate rate of location.

关 键 词：智能交通系统 车牌二次定位 改进算法 定位准确率 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]