基于计算机视觉的接触轨检测车振动补偿方法及应用

Computer Vision Approach to Conductor Rail Detecting Vehicle Vibration Compensation Method and Its Application

以轨道为参考建立世界坐标系,以车体为参考建立摄像机坐标系。考虑摄像机镜头畸变,建立摄像机非线性模型,采用基于最小二乘法的共面标定方法标定摄像机,求解摄像机参数矩阵。根据车体振动特性,推导基于计算机视觉技术的车体振动偏移补偿计算方法,并将该补偿方法成功运用于接触轨检测车。通过广州地铁四号线区间运行数据,验证了该补偿方法能够有效提高系统检测精度。

The world coordinate system was constructed with tracks as the reference system, and the camera co- ordinate system was constructed with vehicle bodies as the reference system. Taking into consideration lens distortion, the non-linear model of cameras was set up and the cameras were calibrated with the coplanar meth- od based on the least square method to acquire the camera parameters matrix. According to vehicle body vibra- tion characteristics, the method to calculate vehicle body vibration displacement compensations was deduced on the basis of computer vision technology and was successfully applied in the conductor rail detecting vehicle. The data detected in the Chebeinan-Huangcun Section of Guangzhou Metro Line 4 demonstrate the proposed com- pensating method can effectively improve accuracy of system detection.