边缘特性筛选与多判定机制下的车牌定位方法  

Approach for license plate location using edge features filter and multi-decision mechanism

作　　者：马爽[1] 陈江宁[2] 卢虎[3] 樊养余[1] 

机构地区：[1]西北工业大学电子信息学院,西安710072 [2]西北工业大学自动化学院,西安710072 [3]空军工程大学理学院,西安710051

出　　处：《计算机工程与应用》2014年第9期145-149,共5页Computer Engineering and Applications

基　　金：国家自然科学基金(No.61174194);航空科学基金(No.20110196004)

摘　　要：针对复杂背景下的车牌定位,利用目标区域的边缘梯度特性筛选出车牌边缘;选取合适的结构元素做数学形态学填充,得到车牌候选区域;融合车牌灰度纹理特征与颜色特征通过多判定机制剔除伪车牌区域,实现车牌的准确定位;通过实验对大量实拍的复杂背景下的车辆图像进行测试。结果表明,该方法准确率高、速度快,克服了传统算法对拍摄环境、光照条件、颜色、角度等较为敏感的问题,具有良好的定位效果。This paper presents a method of license plate location based on edge features filter and multi-decision mecha-nism under complex backgrounds. The edges of license plate are retained by using the edge features of target region. The candidate regions are acquired by the morphological processing. This paper adopts the multi-decision mechanism which combines the texture features and the color features of license plate to remove the pseudo plate regions. Experimental results of many images under complex backgrounds show that this method is effective and fast, which overcomes the sensitivity to image environment, illumination, color, and shooting angle.

关 键 词：车牌定位 边缘特性筛选 数学形态学 灰度纹理特征 颜色特征 多判定机制 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]