基于模板匹配图像算法的高速铁路接触网几何参数测量

摘要

利用机器视觉三角测量原理,建立基于弓网视频的接触网几何参数测量参考坐标系,并进行测量标定.通过合理提取模板类型,利用时序图像滤波方法,实现连续视频图像快速匹配与参数测量.以海南东环铁路公司综合检测列车200 km/h的弓网视频检测数据为例,对比接触式检测系统测量数据和现场实测数据,分析测量结果准确性,用3次测量结果分析重复性.分析结果表明,基于模板匹配方法测量接触网几何参数方案可行,在200 km/h运行速度下,测量误差可控制在±20 mm.%By applying the machine vision triangulation principle,this paper establishes a reference coordinate system for OCS geometric parameter measurement based on pantograph-catenary video to measure and calibrate. By rationally extracting template type and utilizing sequential image filtering, it can realize continuous fast video image matching and parameter measurement. Taking the 200 km/h pantograph-catenary video test data on the comprehensive inspection train of Hainan East RingRailway for example,the test data and the field data on the contact system has been compared to analyze the accuracy of the measurement results,with the repeatability analyzed based on the measurement results of three times. The analysis results show that geometric parameter measurement based on template matching image algorithm is feasible. With a running rate of 200 km/h,the measurement error can be controlled within ±20 mm.