摘要
汽车牌照自动识读系统是机器视觉和模式识别技术在现代智能交通领域中的一项重要应用,其中车牌定位技术是车牌自动识读系统正确识读的前提和关键。本文在分析了现有定位技术的优缺点基础上,提出了一种综合定位算法。该算法通过对视频图像中运动车辆的检测,消除复杂背景的干扰,利用区域生长、阴影消除、彩色分割及形态学分析等图像处理方法,获取车牌的精确定位。最后给出实验结果,实验表明,该算法有效定位率达到92%,具有较好的实用价值。
Car license plate identification system is machine vision and pattern recognition technology in the field of modern intelligent transportation is an important application,the vehicle license location is the key and premise of correct identification.This paper analyzes the advantages and disadvantages of the existing positioning technology based on an integrated location algorithm proposed.The algorithm of moving video image vehicle detection,eliminating the interference of complex background,use of regional growth,shadow elimination,color segmentation and morphological analysis,image processing methods to obtain the exact position of license plate.Finally,experimental results,experiments show that the algorithm is effective 92% localization rate,it has practical value.
出处
《电子测量技术》
2010年第8期35-38,共4页
Electronic Measurement Technology
关键词
视频图像
车牌定位
复杂背景
video image
license plate location
complex background
