摘要
提出在车牌图像定位的预处理中采用顶帽变换,对车牌背景进行估计和消除。创建直径大于16的结构元素,将其分解为几个简单的结构元素,以利于提高灰度形态学运算速度。对车牌图像,顺序采用所得的多个结构元素,先灰度腐蚀后灰度膨胀,即灰度开启运算,得到近似的背景图像。再将原图像减去背景图像,从而消除了背景,保留了车牌区域,利于定位处理。
Top-hat transformation can be used in image preprocessing for location of license plate, in order to estimate and eliminate the background of license plate image. A structure element is established which diameter is more than 16. Then the structure element is split to a few simple structure elements, for the increasing of speed in grey morphologic operation. For license plate image, the several structure elements are adopted in proper order, grey dilation after grey erosion namely grey open operation are ta...
出处
《微计算机信息》
北大核心
2008年第9期310-311,共2页
Control & Automation
基金
2006年国家自然科学基金项目(60572078)
关键词
顶帽变换
背景估计
背景消除
车牌定位
Top-hat transformation
background estimation
background elimination
license plate location
