CAJ | 학술논문

通过多个导引驱动模块可构成用于重载搬运的自动导引车(AGV)，固定在每个模块上的摄像机组成了多目视觉主动导引系统。本文针对多驱动模块的协同导引，提出了一种多摄像机主动导引系统的协同标定方法。首先，通过平面模板法完成单个驱动模块上摄像机内部参数、外部参数的标定；然后通过控制各个驱动模块协同运动，利用特征轨迹约束法估计摄像机相对于驱动模块的位姿误差、不同驱动模块之间的装配误差。建立单个模块的视觉测量坐标系与 AGV 整车运动控制坐标系之间的变换模型。不依赖精密辅助测量工具，具有较高的精度及可实施性。
통과다개도인구동모괴가구성용우중재반운적자동도인차(AGV)，고정재매개모괴상적섭상궤조성료다목시각주동도인계통。본문침대다구동모괴적협동도인，제출료일충다섭상궤주동도인계통적협동표정방법。수선，통과평면모판법완성단개구동모괴상섭상궤내부삼수、외부삼수적표정；연후통과공제각개구동모괴협동운동，이용특정궤적약속법고계섭상궤상대우구동모괴적위자오차、불동구동모괴지간적장배오차。건립단개모괴적시각측량좌표계여 AGV 정차운동공제좌표계지간적변환모형。불의뢰정밀보조측량공구，구유교고적정도급가실시성。
A heavy-duty transport automatic guided vehicle (AGV) can be constituted by multiple guiding driving modules. Cameras fixed on each module compose the initiative multi-camera guiding system. A collaborative calibration method of the multi-camera initiative guiding system is used for the collaborative guiding of multiple driving modules. Firstly, the internal parameters, external parameters and distortion parameters of each camera are calibrated with planar patterns. Then, characteristic trajectory constraint is proposed to estimate the position and orientation error, and the error of the drive module to the vehicle body by controlling the coordinated motion of driving modules. Transformation model from the vision measurement coordinate to vehicle motion control coordinate is constructed, which is the technical foundation for the collaborative guidance of multi-camera initiative guided AGV. The method can realize the calibration without need of the precise auxiliary measuring instruments, and with a higher precision and implement ability.