摘要
提出基于机器视觉的车辆动态偏移量检测方案,即在被测车辆非共线的4个点上安装高速电荷耦合器件(CCD),随着车辆的运动,CCD高速扫描钢轨,实时分析钢轨相对CCD的运动图像序列,得到各测点相对于钢轨的横向位移和垂向位移,综合各测点的相对偏移量得到整个车体的运行姿态,经进一步分析得到任意点的动态偏移量;研究检测原理,建立由各测点的动态偏移量计算描述车体运行姿态的5个变量和计算任意点动态偏移量的公式,研制模拟车体振动的平台,将所研发的系统与第3方地面测试设备对模拟平台动态偏移量的测试结果进行比较。研究结果表明,采用这2种方法测得相同位置的偏移量波形完全相似,对应幅值相差在-2～2mm以内。
The scheme based on machine vision for detecting the offset was put forward. The scheme is as follows： four high-speed charge couple devices are non-collinearly mounted on car-body; with the movement of car-body, CCDs scan the rails, and images are analyzed in real time to get four locations＇ lateral and vertical displacement relative to rail; attitude of the whole car is obtained by analyzing displacement of those locations introduced, and some formulae to calculate five variables describing car-body attitude and random location＇s offset are built. A test-rig platform which can simulate vehicle＇s vibration was developed. Test results obtained respectively by this system and other test instrument fixed on ground were compared. The results show that waveforms obtained by those two system are similar and corresponding amplitudes only differs within 2 mm.
出处
《中南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2008年第4期787-792,共6页
Journal of Central South University:Science and Technology
基金
铁道部科技研究开发计划资助项目(Z2007-079)
关键词
机器视觉
动态偏移量
车辆运行姿态
在线检测
machine vision
offset generated by vibration
attitude of vehicle
on-line detection
作者简介
通信作者：韩伯领（1964-），男，江苏淮安人，博士研究生，讲师，高级工程师，从事铁路运输研究；电话：0731-2655294；E-mail：qlzjzd@mail．csu．edu．cn
