摘要
该文提出一种基于数学形态学的车牌图像分割算法。首先,对原始车牌图像进行预处理;然后,利用一种抗噪型数学形态学边缘检测算子进行车牌边缘检测;最后,去除图像水平噪声,结合水平、垂直投影进行定位。实验表明,该方法不仅算法简单、准确度高,且适于对有噪声及背景复杂的车牌图像进行分割。
In this paper, a algorithm of license plate segmentation based on mathematical morphology is proposed. First, we carry on the pretreatment to the original license plate image; Then,detect the car licence edge with one anti-chirp mathematics morphology edge detection operator; Finally, remove the horizontal noise of car licence mage and union horizontal, vertical positioning projection for locating the car licence. During the experiment, This method is not only a simple algorithm、high accuracy, and suitable for license plate segmentation to the car license image which have the noise and complex background.
出处
《电脑知识与技术》
2009年第8X期6859-6861,共3页
Computer Knowledge and Technology
关键词
车牌分割
数学形态学
结构元素
边缘检测
license plate segmentation
mathematical morphology
structure element
edge detection
