基于不变矩的机器视觉车辆类型识别技术  

Research on Vehicle Type Recognition by Computer Vision Based on Invariant Moments

作　　者：翟乃斌[1] 苏建[1] 宋年秀[2] 刘玉梅[1] 陈友谊[1] 

机构地区：[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

出　　处：《武汉理工大学学报（信息与管理工程版）》2007年第4期7-10,共4页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：吉林省自然科学基金资助项目(20030531-1);教育部博士点基金资助项目(20030183032)

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。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 ex- traction 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.

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 

分 类 号：TP319.41[自动化与计算机技术—计算机软件与理论] U467.5[自动化与计算机技术—计算机科学与技术]