灰度图像下车牌定位与分割算法研究  

Study on the Location and Segmentation Algorithm of Vehicle License Plate with Gray Image

作　　者：刘培锋[1] 张文斌[1] 王祁[1] 

机构地区：[1]哈尔滨工业大学电气工程及自动化学院,150001

出　　处：《微计算机信息》2008年第18期281-283,共3页Control & Automation

基　　金：黑龙江省交通(HJZ_2004_12)项目名称:重载车辆对水泥混凝土路面破坏作用的研究

摘　　要：给出了一种能在复杂交通环境下对汽车牌照进行定位和分割的算法。算法都是基于灰度图像的。在定位算法中,首先对图像进行边缘检测和水平膨胀,然后进行连通区域分析、水平聚类,最后对得到的车牌候选区域进行模糊决策,找出真正车牌区域,很好的解决了伪车牌问题。在分割算法中,利用最小二乘法对车牌进行倾斜校正,利用可变小模板对字符进行分割。实验证明,定位准确率为97%,分割准确率为95.3%,具有非常好的效果。An algorithm to obtain location and segmentation of vehicle license plate under complicated traffic background is provided. The algorithm is based on gray image. In the study on this localization algorithm, firstly image edge detection and horizontal dilate are carried out, then connected area analysis and horizontal clustering are presented, finally in order to better solved the problem of false vehicle plate, the fuzzy decision-making is carried on the obtained vehicle license area. In the study on segmentation, least squares method is used for tilt correction, and variable small template is used for character segmentation. Experimental results show that the method is quite effective, and the plate location rate of success is 97.8%, the character segmentation rate of success is 95.3%.

关 键 词：连通区域 水平聚类 模糊决策 最小二乘法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]