无人地面平台多传感器的联合外标定方法  

Integration Extrinsic Calibration Algorithm of Unmanned Ground Platform Multi-Sensor

作　　者：骆云志 张春华[1] 吕卫强 周建平[1] 王钤[1] 赵永滨 李健[1] 李珏颖[1] 

机构地区：[1]中国兵器工业第五八研究所军品部,四川绵阳621000

出　　处：《兵工自动化》2012年第12期66-75,80,共11页Ordnance Industry Automation

摘　　要：针对现有的传感器标定方法的不足,提出一种采用3D标定物的无人地面平台多传感器联合外标定方法。在参考和借鉴国内外学者的智能移动机器人传感器标定研究成果的基础上,建立了无人地面平台和多传感器坐标系;并从3D标定物、多层激光测距仪的外标定、多层激光测距仪和CCD摄像机的联合外标定、坐标系之间的齐次变换和L-M非线性最优算法等几个方面论述多传感器的外标定方法和步骤;对标定结果进行了验证和误差分析。现该方法已经成功应用于中型无人地面平台项目。应用结果表明:该方法能修正各传感器的安装位姿误差,保证标定后的各传感器的测量精度和目标(障碍物)在平台坐标系的一致性。Aiming at the shortcoming of current sensor calibration algorithm, introduce the integration extrinsic calibration algorithm of unmanned ground platform multi-sensor based on 3D calibration item. Based on research of foreign and demotic intelligence mobile robot sensor calibration, establish the unmanned ground platform and multi-sensor coordinate system. Introduce the multi-sensor extrinsic calibration method and its steps based on 3D calibration item, extrinsic calibration of multi-layer laser range finder and CCD camera, homogeneous transformation among coordinate system, and L-M nonlinear optimum algorithm. Carry out validation and error analysis of calibration. The method is successfully used in middle-size unmanned ground platform project. The application results show that the method can correct the installation position error of each sensor, and ensure the consistency of calibrated sensor measurement accuracy and object （obstacle） in platform coordinates system.

关 键 词：无人地面平台 多层激光测距仪外标定 多传感器联合外标定 L—M算法 

分 类 号：TJ02[兵器科学与技术—兵器发射理论与技术]