论文部分内容阅读
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。
In order to solve the problem of space synchronization of intelligent vehicle information fusion system using radar and machine vision, a four-point calibration method of space synchronization between two sensors is proposed. Based on the relationship between the two sensors’ coordinate system, this method selects the 10 m area in front of the vehicle as the calibration area. Two pairs of point pairs on two lines parallel to the central axis of the radar are used in this method. The horizontal angle of the longitudinal axis of the sensor coordinate system and the camera’s pitch angle are calibrated. Static verification tests and dynamic verification tests show that this method can obtain high spatial synchronization accuracy of the obstacle distance data acquired by the two sensors and can be calibrated using only 4 points in the area of 10 m in front of the vehicle, Easy to do.