基于先验知识的红外图像汽车牌照定位方法  

Infrared Image Car License Plate Location Method Based on Apriori Knowledge

作　　者：胡爱明[1] 周孝宽[1] 

机构地区：[1]北京航空航天大学宇航学院

出　　处：《北京航空航天大学学报》2003年第2期128-131,共4页Journal of Beijing University of Aeronautics and Astronautics

摘　　要：由于汽车牌照图像背景复杂 ,存在各种各样的噪音干扰 ,很难准确定位汽车牌照 .因此提出一种快速、自动红外汽车牌照定位方法 ,利用红外成像特点 ,解决了高速行驶、无主动照明、在全天候工作情况下的牌照定位问题 .根据汽车牌照在全图单位长方形面积的水平灰度梯度最大的特点 ,成功地从红外汽车图像中检测定位车牌 ,定位结果的正确率达到了Since the background of a car license plate in an infrared image is quite complex, so it is difficult to locate it. A fast infrared image car license plate automatic location system is presented. The system is suitable for locating the position of car license plate in an infrared image under all-weather condition. Based on the feature of a car license plate image possessing of the highest average gradient in whole image, location system is designed and implemented, and the accuracy of location rate can reach 99%.

关 键 词：先验知识 红外图像 汽车牌照 字符识别 智能交通 

分 类 号：U462.2[机械工程—车辆工程] TP391[交通运输工程—载运工具运用工程]