一种基于TOF相机与CCD相机的联合标定算法研究

Joint calibration technique for TOF camera and CCD camera

针对基于time-of-flight(TOF)相机的彩色目标三维重建需标定CCD相机与TOF相机联合系统的几何参数,在研究现有的基于彩色图像和TOF深度图像标定算法的基础上,提出了一种基于平面棋盘模板的标定方法。拍摄了固定在平面标定模板上的彩色棋盘图案在不同角度下的彩色图像和振幅图像,改进了Harris角点提取,根据棋盘格上角点与虚拟像点的共轭关系,建立了相机标定系统模型,利用Levenberg-Marquardt算法求解,进行了标定实验。获取了TOF与CCD相机内的参数,并利用像平面之间的位姿关系估计两相机坐标系的相对姿态,最后进行联合优化,获取了相机之间的旋转矩阵与平移向量。实验结果表明,提出的算法优化了求解过程,提高了标定效率,能够获得较高的精度。

The geometric parameters of CCD camera and TOF camera system is need to know first in color target 3D reconstruction based on time-of-flight(TOF)camera,this paper proposed a calibration method based on planar pattern,after researching the calibration algorithm of color image and TOF depth image existed.It took color images and amplitude images from different angles of the checkerboard pattern,which fixed on a flat calibration template,improved Harris corner extraction algorithm.It established the model of camera calibration system according to the conjugate relation between the corner point and the virtual image point and used Levenberg-Marquardt algorithm to solve,then carried out calibration experiments.Obtaining the internal parameters of the TOF and CCD cameras,it estimated the relative attitude of the two camera coordinate system by using the pose relation between image planes.Finally,it performed joint optimization and obtained the rotation matrix and the translation vector between the cameras.Results show that the proposed algorithm optimizes the solving process,improves the efficiency of calibration,and obtaines higher accuracy.