基于手眼模型的三维激光雷达外参数标定

发布时间：2018-03-01 05:15

本文关键词： 无人车三维激光雷达点云配准手眼模型外参数标定　出处：《光电工程》2017年08期 　论文类型：期刊论文

【摘要】：无人车三维激光雷达与GPS/INS组合导航系统融合使用,需要对两者之间的相对位姿进行标定。针对车辆运动过程点云畸变的现象,提出了一种单帧点云中激光扫描点坐标修正方法。针对标定问题,建立了标定模型,提出了一种基于ICP算法和手眼标定模型及最小二乘法进行外参数标定的方法。通过蒙特卡罗仿真实验,对该方法的有效性和精度进行了仿真和验证。在实验室JJUV-6无人车平台上进行标定试验,得到两者之间的三维位姿关系。对比标定前后三维点云重建效果,点云重叠度接近于配准效果。基于应用的目的,基本满足二维地图构建和三维环境重建的需求。
[Abstract]:It is necessary to calibrate the relative position and attitude between the three dimensional lidar and GPS/INS integrated navigation system of unmanned vehicle. In this paper, a method for correcting the coordinates of laser scanning points in a single frame point cloud is proposed. A method of external parameter calibration based on ICP algorithm, hand-eye calibration model and least square method is proposed. The validity and accuracy of the method are simulated and verified. The calibration experiments are carried out on the JJUV-6 unmanned vehicle platform in laboratory, and the three dimensional position and pose relationship between the two methods is obtained. The effect of 3D point cloud reconstruction before and after calibration is compared. The overlap degree of point cloud is close to the registration effect. For the purpose of application, it can basically meet the needs of 2D map construction and 3D environment reconstruction.
【作者单位】：军事交通学院;
【基金】：国家自然科学基金重大项目(91220301) 国家重大研发计划(2016ffi b0100903)
【分类号】：TN958.98

【相似文献】