基于区域标注的车牌提取算法研究

提出了一种基于区域标注的方法来实现静态图像中的车牌提取。车牌提取的主要步骤由以下几步构成:原始图像灰度化、数学形态学处理(前景图像处理)、图像二值化(阈值法)、图像滤波处理、特征区域标注,车牌特征提取、采用matlab进行实验仿真。在实验中,通过对各种场景下采集到1 000幅图像做测试,车牌定位率为98%。实验结果表明基于区域标注的车牌提取算法运算快,对背景较为复杂的图像中车牌的提取有良好的效果。

Study of Licence Plate Extraction Algorithm based on Region Labeled

This paper introduces the method of licence plate extraction in static image based on region labeled.The process of licence plate extraction mainly include:gray of orginal image,mathematical morphology processing(foreground image processing),image binarization(threshold approach), image filtering, binarization,feature region labeled, Licen extraction.We used MATLAB to achieve this simulation of processing.After 1000 images from the real sence is tested,the accuracy of locating the licence plates is 98%.The result of experiment illustrates that this algorithm based on region labeled have a high speed,and the extraction is accurate for the image of which background is complicate.