激光雷达与相机联合标定进展研究
Review of joint calibration of LiDAR and camera
熊超 1乌萌 1刘宗毅 1卢传芳2
作者信息
- 1. 地理信息工程国家重点实验室,西安 710054;西安测绘研究所,西安 710054
- 2. 西安航天天绘数据技术有限公司,西安 710100
- 折叠
摘要
针对激光雷达与相机标定精度会极大地影响定位初始化,融合定位算法对标定参数的准确性非常敏感,会严重降低融合定位算法的性能和可靠性等问题,提出一种激光雷达与相机联合标定分类方法:论述激光雷达与相机联合标定领域的最新的研究进展;重点分析和总结基于特征的标定方法、基于运动的标定方法、基于互信息的标定方法和基于深度学习的标定方法的突出研究成果;总结出4种标定方法的开源代码工具集,并对这4种标定方法的特点、标定精度和自动化程度进行分析比较;最后,展望激光雷达与相机联合标定研究的发展趋势.
Abstract
Aiming at the problems that the calibration accuracy of light detection and ranging (LiDAR) and camera can greatly affect fusion positioning initialization,and localization algorithms are very sensitive to the accuracy of calibration parameters,which may seriously reduce the performance and reliability of the localization algorithm,the paper proposed a classification method of combined calibration of LiDAR and camera:the latest developments in this field were discussed;and the highlight research achievements of feature-based method,the motion based method,the mutual information based method and deep learning based method were analyzed and summarized emphatically;then,the open source code tools of those four methods were listed,and their features,calibration accuracy and automation level were comparatively analyzed;finally,the future development trends of combined calibration of LiDAR and camera were concluded.
关键词
激光雷达(LiDAR)/相机/双目视觉/棋盘格/外参标定/联合标定Key words
light detection and ranging (LiDAR)/camera/stereo vision/checkerboard/extrinsic calibration/joint calibration引用本文复制引用
基金项目
地理信息工程国家重点实验室研究课题(SKLGIE2022-ZZ-03)
出版年
2024
NETL
NSTL
万方数据