基于ICP配准的激光雷达相机联合标定算法

A Joint Calibration Algorithm for LiDAR Cameras Based on ICP Registration

激光雷达与相机联合标定是实现两者采集数据高精度融合的基础,传统的联合标定方法多采用特征点识别的方式,易受环境噪声及激光雷达分辨率等问题的影响。文章提出一种基于ICP配准的联合标定算法,采用设计的标定物模板,通过ICP配准算法实现实际标定物点云与标定物模板点云间精确对齐,并根据标定物模板特定角点位置,获取实际标定物点云角点坐标。在此基础上,通过PnP算法求解激光雷达与相机间转换矩阵,进行激光雷达与相机所对应点云坐标与像素坐标间的转换,实现两传感器之间的高精度联合标定。试验结果表明,文章提出的联合标定算法能够有效提高激光雷达与相机间的联合标定精度,可为更准确的目标检测和跟踪提供高精度联合测量基础。

The joint calibration of LiDAR and camera is the basis for achieving high-precision fusion of collected data.Traditional joint calibration methods often use feature point recognition,which is susceptible to environmental noise and LiDAR resolution issues.This article proposes a joint calibration algorithm based on ICP registration,which uses a designed calibration object template to achieve precise alignment between the actual calibration object point cloud and the calibration object template point cloud through the ICP registration algorithm.Based on the specific corner position of the calibration object template,the actual calibration object point cloud corner coordinates are obtained.On this basis,the conversion matrix between the LiDAR and the camera is solved using the PnP algorithm to convert the point cloud coordinates and pixel coordinates corresponding to the LiDAR and the camera,achieving high-precision joint calibration between the two sensors.The experimental results show that the proposed joint calibration algorithm can effectively improve the joint calibration accuracy between LiDAR and camera,providing a high-precision joint measurement foundation for more accurate target detection and tracking.