一种自适应摄像机与激光雷达联合标定算法  

An Adaptive Camera and Lidar Joint Calibration Algorithm

作　　者：姚文韬[1] 沈春锋[1] 董文生[1] 

机构地区：[1]上海宝信软件股份有限公司,上海201203

出　　处：《控制工程》2017年第S1期75-79,共5页Control Engineering of China

基　　金：上海市科委基础研究重大重点项目:面向工业大型机械智能网络控制的多信息视觉识别和三维重构关键技术研究(14JC1402203)

摘　　要：无人自主车依赖于摄像机与激光雷达对周围环境进行感知。为了对两者的信息进行融合,首先要通过两者的联合标定将数据置于同一坐标系下。针对摄像机与三维激光雷达的联合标定问题,提出了一种自适应的标定方法。仅需一块平面标定板即可完成全部标定过程。在非线性优化过程中,通过交替优化激光雷达至摄像机坐标系的旋转矩阵及平移向量,以及摄像机外参数,相比固定摄像机外参数的方法,平均标定误差可减小22.50%。通过对每帧激光雷达数据进行加权,平均标定误差可进一步减小8.80%。Unmanned autonomous vehicles rely on camera and lidar to percept the environment. In order to fuse the information of the camera and lidar, joint calibration is essential to place their data under the same coordinate system. To solve the problem of camera and 3 D lidar joint calibration, an adaptive joint calibration method is proposed. Only one planar calibration board is needed for the calibration. In the nonlinear optimization procedure, the camera extrinsic parameters are optimized together with the rotation matrix and translation vector from lidar to camera coordinate. Compared with the method of fixed camera parameters, the average calibration error is reduced by 22.50 %. By using the weighed lidar data, the calibration error is further reduced by 8.80 % on average.

关 键 词：摄像机 激光雷达 联合标定 信息融合 无人自主车 

分 类 号：TN958.98[电子电信—信号与信息处理] TP391.41[电子电信—信息与通信工程]