摘要
基于多种格式车牌的共性,提取车牌图像的纵向边缘,然后根据车牌区边缘灰度跳变和边缘密度等特征,采用一系列步骤去除无效和干扰边缘,保留类车牌特征区域;通过横向形态学运算,使类车牌区闭合,有效克服了形态学结构元素难以随车牌大小变化自适应选取的问题.基于扫描线种子填充算法来搜索车牌区域,利用颜色信息进行反色判断,并基于边缘检测的方法来进行车牌区域二值化,利用形态学和连通域检测方法进行字符的精细切分.对实际场景中大量车牌样本加以验证表明,算法准确率高,对车牌大小自适应性良好,具有较好实用价值.
Based on the common features of a variety of license plates, the vertical edge was first detected. Then, some approaches were adopted to remove the invalid edge due to the characteristics of edge gray level jump and edge density, so that the regions having features of license plate were preserved. Next, horizontal mathematical morphology (MM) operation was conducted to close the candidate regions, and the above method overcomes the problem how the MM structure element was adjusted to the size change of license plates. By scan-line seed fill algorithm, the license plate region was searched. Then, color-reversing judgement was conducted by color, and binarization was done based on edge detection. Afterward, characters were segmented on the basis of MM and connected components analysis. With an abundant samples verified in dark hours and daytime in real conditions, the experiment indicates that it is feasible to adopt this algorithm in license plate recognition system to achieve accuracy and adaptability.
出处
《微电子学与计算机》
CSCD
北大核心
2009年第4期217-221,共5页
Microelectronics & Computer
基金
教育部跨世纪优秀人才培养计划基金项目(2003714)
关键词
车牌识别
车牌定位
字符分割
图像处理
边缘特征
自适应性
license plate recognition
license plate orientation
character segmentation
image processing
edge characteristic
adaptability