基于阈值分割的运矿车辆车牌精确定位  

Accurate license plate location for mineral transport vehicle based on threshold segmentation

作　　者：胡金蓉[1] 王玲[1] 

机构地区：[1]四川师范大学计算机科学学院,四川成都610068

出　　处：《计算机工程与设计》2009年第8期2051-2054,共4页Computer Engineering and Design

摘　　要：由于运矿车辆车牌图像质量低、车牌污染严重、环境复杂、识别率低等特殊情况,设计了一种基于灰度阈值分割的运矿车辆车牌的精确定位算法。它通过对粗定位车牌图像进行灰度阈值分割,得到估计车牌区域并进行区域标记,根据车牌所在连通区域面积最大这一特征,得到完整有效的车牌区域,结合车牌先验知识去掉其边框,从而得到车牌的精确定位结果。实验结果表明,该算法定位准确率高、耗时少、分割效果好,具有较强的实用价值,是对现有车牌识别技术有意义的扩展和补充。An accurate license plate location algorithm for mineral transport vehicle based on gray threshold segmentation is designed, because of the special circumstances ofmineral transport vehicle such as low license plate image quality, serious pollution vehicles, complex environment, low recognition rate and so on. Firstly, the estimate plate region is found through gray threshold segmentation for the rough positioning plate image and made area labeling to it. Secondly, the complete and correct license plate region is found according to the feature that the plate ofconnected region＇s area is the largest. Finally, the accurate plate region is found by erasing the plate＇s border according to its prior knowledge. Experimental results demonstrate it is an effective, less time-consuming and strong practical valuable ac- curate license plate location algorithm. It would be a potentially significant contribution to the active area of license plate recognition system.

关 键 词：运矿车辆 车牌定位 阈值分割 车牌边框 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]