摘要
在智能车牌识别系统中,经过汽车图像的初分割,可从复杂的汽车及背景图像中分割出汽车图像,但复杂灰度图像经二值化处理后存在许多干扰白点.提出一种“高斯滤波”的方法去除这些白点,使二值化图像变得平滑、少毛刺,并具有较好的连通性.
In the intelligent recognition system of car license plate, car image can be separated from complex car background image by preliminary separation of car image ,but in these threshold images, the complex gray image elements are translated into some white noise. This paper presents a Gauss filter method to eliminate noise of threshold images, the threshold image is changed into smooth one which has smaller burr and has better connectedness.
出处
《武汉理工大学学报(交通科学与工程版)》
2005年第5期805-806,共2页
Journal of Wuhan University of Technology(Transportation Science & Engineering)
基金
国家民委重点科研项目资助(批准号:Mzy02004)