Abstract: In this paper，a new algorithm which can improve the accuracy of vehicle license plate location using Top-hat transform is presented.Based on texture feature of license plates and the characteristic of Top-hat transform，the Top-hat coefficients are retained.The license plate can be located through smoothing，vertical direction information，four connected region detection，density calculation and so on.The proposed algorithm is suitable for the single-license and multi-license plate detection in the various illumination conditions with high accuracy.The performance of the proposed algorithm has been tested on monitoring images obtained from the practical application.Experimental results show that this algorithm has better performance than others’ on location precision，misdetection rate，and misjudgment rate.

Key words: image processing, Top-hat transform, morphology, vehicle license plate localization

摘要： 为了提高多车牌定位的精度，提出了一种基于Top-hat变换的车牌定位算法。该算法根据车牌纹理的空间分布特点和Top-hat变换的特点，以Top-hat系数为依据，通过平滑、二值化、垂直信息提取、连通、密度计算等多个操作完成车牌定位，该算法可对单、多车牌定位。以实际监控的图像为实验数据，通过对比实验，其结果表明：在不同背景情况下，方法在定位精度、漏判率和误判率等方面均优于其他算法。

关键词: 图像处理, Top-hat变换, 形态学, 车牌定位