一种基于视频流的多车牌定位方法  

Approach to Locate Multi-License Plates Based on Video Streams

作　　者：董春雨[1] 王夏黎[1] 刘炜琪[1] 谢生龙 

机构地区：[1]长安大学信息工程学院,西安710064

出　　处：《计算机系统应用》2014年第9期230-235,共6页Computer Systems & Applications

基　　金：国家自然科学基金(50978030);中国博士后科学基金(2012M521729)

摘　　要：针对视频流中多个车牌的定位问题,提出了一种基于图像复原技术与HSV颜色空间的多车牌定位方法.该方法的主要思想是用改进的维纳滤波器对获取的视频图像进行复原,根据车牌的颜色特征,利用HSV颜色模型构造5级灰度图,对灰度图用形态学方法进行消噪和区域连通处理,最后利用水平、垂直投影技术,结合车牌先验知识,通过分析和判断来确定并分割出各个汽车牌照.实验表明,该方法能有效的从复杂的背景中准确定位并提取出多个运动中的车牌.For the location of multiple License Plates in video streams, this paper presented an effective method of Multi-license plates location based on image restoration and HSV color model. The main idea of this method was to use an improved Wiener filter to get the video image recovery. And then in HSV color space, it constructed a five-scale gray image according to the color characteristics of license plate, and used the morphology smooth to eliminate noise. Finally, the potential region of the plate was segmented through horizontal and vertical projection value and we got the accurate position of the multiple plates by prior knowledge of the car-plate region. Experimental results show that the proposed method can extract multiple motive license plates exactly from a complex background.

关 键 词：视频流 多车牌定位 图像复原 HSV颜色空间 形态学方法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]