首页 | 本学科首页   官方微博 | 高级检索  
     

基于不变矩的机器视觉车辆类型识别技术
引用本文:翟乃斌,苏建,宋年秀,刘玉梅,陈友谊. 基于不变矩的机器视觉车辆类型识别技术[J]. 武汉理工大学学报(信息与管理工程版), 2007, 29(4): 7-10
作者姓名:翟乃斌  苏建  宋年秀  刘玉梅  陈友谊
作者单位:1. 吉林大学,交通学院,吉林,长春,130022
2. 青岛理工大学,汽车与交通学院,山东,青岛,266033
基金项目:吉林省自然科学基金;高等学校博士学科点专项科研项目
摘    要:针对智能交通领域汽车类型识别的应用背景,利用单目视觉开发了基于计算机视觉车辆类型识别系统;论述了单目视觉原理及特征提取的关键技术,利用不变矩及不变矩矢量在图像平移、旋转及比例变换时保持不变的特性,以其作为主要特征实现车辆类型有效识别。试验结果表明,该技术可以实现车辆类型的自动、快速和准确分类。

关 键 词:智能交通  计算机视觉  不变矩  不变矩矢量  车辆类型识别
文章编号:1007-144X(2007)04-0007-04
修稿时间:2006-11-28

Research on Vehicle Type Recognition by Computer Vision Based on Invariant Moments
ZHAI Naibin,SU Jian,SONG Nianxu,LIU Yumei,CHEN Youyi. Research on Vehicle Type Recognition by Computer Vision Based on Invariant Moments[J]. Journal of Wuhan University of Technology(Information & Management Engineering), 2007, 29(4): 7-10
Authors:ZHAI Naibin  SU Jian  SONG Nianxu  LIU Yumei  CHEN Youyi
Abstract:Based on the application of vehicle type recognition in intelligent traffic fields,a vehicle type recognition system based on computer was developed,adopting the technique of monocular vision.The key techniques of monocular vision and feature extraction were discussed.The recognition of vehicle type is achieved by applying invariant moment and invariant moment vector,which keep invariant when the image is in different position,orientation and scale.It has been tested that the system can recognize vehicle types automatically and accurately in a quick way.
Keywords:intelligent traffic  computer vision  invariant moments  invariant moment vector  vehicle type recognition
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号