飞行时间深度相机和彩色相机的联合标定

飞行时间（TOF）深度相机等深度传感器可以实时、准确地获取深度信息，在计算机视觉领域受到广泛关注。本文以获取同一视野下场景的纹理信息和深度信息为目的，在传统的棋盘格标定方法基础上，针对TOF深度相机的低分辨率和较大的径向畸变的特点，采用角点稀疏的棋盘格作为标定板以提高角点检测的精度，提出一种TOF深度相机和彩色相机的联合标定方法。首先对彩色相机和深度相机单独标定，使用传统的棋盘格标定方法获得彩色相机的内部参数，对使用深度相机所拍的强度图进行畸变校正后再求得深度相机的内部参数。然后，固定两相机内参，多次实验获得两相机之间的相对位置关系，并使用最小二乘法进行优化。实验结果表明，该方法标定精度高，彩色相机的重投影误差最多可减小0.15个像素，深度相机的重投影误差最多可减少0.09个像素，根据标定结果将深度相机所得的深度图投影到彩色相机的视野下所得的投影深度图能和彩色图像精确对齐。 

Joint Calibration of Time-of-Flight Depth Camera and Color Camera

Depth sensor such as Time-of-Flight (TOF) depth camera can get depth information in real time and accurately, which attracts the extensive attention in the computer vision field. To gain the color and depth information of the same scene, a joint calibration method of TOF depth camera and color camera based on traditional chessboard calibration methodis proposed, aiming at the low resolution and high radial distortion of the TOF camera, using the chessboard planar with sparse corners to improve the precision of the corner detection. First, calibrate the color and TOF depth camera separately, the intrinsic parameters of the color camera was gained through the traditional chessboard calibration method, the undistorted intensity images from TOF camera were used to get the intrinsic parameters of the TOF camera. Then, several experiments were conducted to get the relative position of the two cameras, the final result was the value after the least square optimization. Experiments show that the precision of the calibration is high, which means the re projection error of color camera can be reduced 0.15 pixels mostly and error of depth sensor can be reduced 0.09 pixels and the depth image at the color image position acquired by the calibration parameters and the original depth image can be well aligned with the color image.