基于机器视觉的接触网几何参数检测方法

摘要

提出了一种基于机器视觉技术的接触网几何参数检测方法。检测过程中，通过线阵相机对锚段关节、支柱等关键零部件的识别，实现检测数据精确定位，同时采用振动补偿技术消除由检测车体振动所造成的检测误差。检测结果表明，该方法能够有效地提高接触网动态几何参数检测精度，具有实际意义。%A method is proposed based on machine vision technology for catenary geometrical pa-rameters detection.The test data precise positioning is realized by the identification of the ancho-ring section,pillars and other key components.And errors caused by the detector body vibration are eliminated by vibration compensation technology.Results show that this method can effective-ly improve the dynamic geometry detection accuracy.