雷达与机器视觉的空间同步方法

为解决应用雷达与机器视觉的智能车信息融合系统的空间同步问题,提出了对两种传感器进行空间同步的四点标定方法。在建立了两种传感器的坐标系关系的基础上,该方法选取汽车前方10 m区域为标定区域,利用该区域地面上平行于雷达中心轴的两条直线上的两组点对,对两种传感器坐标系纵轴的水平夹角以及摄像机的俯仰角进行标定。静态验证试验和动态验证试验证明,该方法可以获得较高的由两种传感器获取的障碍物距离数据的空间同步准确度,并且只需利用汽车前方10 m区域内的4个点进行标定,简单易行。

Spatial synchronization method for lidar and computer vision

A four points calibration method was developed to eliminate the problem of spatial synchronization of lidar and computer vision systems in the sensor fusion system of intelligent vehicles.A calibration area was selected 10 m in front of the vehicle based on the relationship between the two sensors' coordinates.The yaw angle between the longitudinal axes of the two sensors and the camera pitch angle are calibrated using two pairs of points on two lines which are parallel to the central axis of the lidar.Static and dynamic validation tests show that the method provides accurate spatial synchronization of the two sensors.The method is easy to apply because only four points in front of the vehicle are needed for the calibration.