分层次车牌定位方法  

A Layered Algorithm for Vehicle License Plate Locating

作　　者：吴进军[1] 杜树新[1] 

机构地区：[1]浙江大学工业控制技术国家重点实验室浙江大学智能系统与决策研究所,杭州310027

出　　处：《科技通报》2005年第5期606-611,共6页Bulletin of Science and Technology

摘　　要：车牌定位是车牌识别中的重要步骤,针对汽车图像的复杂背景和不同环境光条件,提出了一种分层次的车牌定位方法。该方法将图像进行分层次分割,不断减少分析的范围。它先定位出车牌候选区域的上下和左右边界,得到多个车牌候选区域,再从中挑选出车牌区域,实现了车牌的准确定位。实验结果表明,该方法定位准确率高,运行速度快,具有很强的鲁棒性。The location of the vehicle license plate is one of the important steps of the recognition of the vehicle license plate. Under the condition of the complicated backgrounds of the vehicle images and the different illumination conditions, a layered location algorithm for vehicle license plate is proposed. In the algorithm the layered gradation segmentation is adopted and the analysis range is diminished gradually in the image. The right-and-left and down-and-left boundaries of plate region candidates are located and a few plate region candidates could be obtained. And then the plate region will be selected from them. So the exact location of the vehicle license plate is carried out. The experimental results show that the algorithm is excellent in the accuracy with a rapid speed and is great in the robustness.

关 键 词：车牌定位 分层次分割 边界定位 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]