基于 ADAS 平台的下线标定系统设计与实现
DOI:
CSTR:
作者:
作者单位:

上海电力大学

作者简介:

通讯作者:

中图分类号:

基金项目:


Design and Implementation of Factory Calibration System Based on ADAS Platform
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    标定是汽车驾驶辅助系统中至关重要的一环,为了使驾驶辅助系统更加精确地感知目标,提升汽车驾驶主动安全性,设计了一种针对摄像头和毫米波雷达的多传感器下线标定系统;该系统通过利用摄像头和毫米波雷达自身结构特性建立数学模型,计算出各传感器的外部参数,将不同传感器感知的目标数据转换到一个统一的世界坐标系下,实现了摄像头和毫米波雷达的单独标定和联合标定;采用了基于CAN总线的UDS诊断服务作为通信媒介,实现了会话控制、安全访问、待机握手、例行程序控制等诊断服务,设计了符合前装量产标准的下线标定流程;为方便产线工人操作和简化前装量产中的操作步骤,设计了专门针对前装量产场景下的上位机标定软件;经过实验测试验证了该下线标定系统的可靠性、实时性和准确性,标定之后的测距误差均值在0.25m附近,毫米波雷达投影点与图像检测框坐标匹配的比率是93.2%,可以满足汽车辅助驾驶系统中摄像头和毫米波雷达感知精度的要求。

    Abstract:

    Calibration is a crucial part of the car driving assistance system. In order to make the driving assistance system perceive the target more accurately and improve the active safety of car driving, a multi-sensor Factory calibration system for cameras and millimeter-wave radars is designed. The system establishes a mathematical model by using the structural characteristics of the camera and the millimeter-wave radar, calculates the external parameters of each sensor, and converts the target data perceived by different sensors into a unified world coordinate system, realizing the camera and the millimeter-wave radar joint calibration. The Unified Diagnostic Services(UDS) based on CAN bus is used as the communication medium, which realizes the diagnostic services such as session control, security access, standby handshake, routine program control, etc., and designs the factory calibration process that conforms to the standard of pre-installation and mass production. In order to facilitate the operation of the production line workers and simplify the operation steps in the front-loading mass production, the upper computer calibration software specially designed for the pre-installation production is designed. The reliability, real-time performance and accuracy of the factory calibration system have been verified by experimental tests. The average value of ranging error after calibration is around 0.25m, and the ratio of the millimeter-wave radar projection point to the coordinates of the image detection frame is 93.2%. The calibration system can meet the perception accuracy requirements of cameras and millimeter-wave radars in automotive assisted driving systems.

    参考文献
    相似文献
    引证文献
引用本文

仲正,宋小军.基于 ADAS 平台的下线标定系统设计与实现计算机测量与控制[J].,2022,30(10):148-154.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2022-05-18
  • 最后修改日期:2022-06-10
  • 录用日期:2022-06-10
  • 在线发布日期: 2022-11-01
  • 出版日期:
文章二维码