一种自动、快速的Kinect标定方法

J4 ›› 2016, Vol. 38 ›› Issue (06): 1193-1199.

一种自动、快速的Kinect标定方法

孟勃，刘雪君

（东北电力大学信息工程学院，吉林吉林 1320122）

收稿日期:2015-07-01 修回日期:2015-09-01 出版日期:2016-06-25 发布日期:2016-06-25
基金资助:
国家自然科学基金（61172111）

A quick autocalibration method of Kinect

MENG Bo,LIU Xuejun

（Academy of Information Engineering,Northeast Dianli University,Jilin 132012,China）

Received:2015-07-01 Revised:2015-09-01 Online:2016-06-25 Published:2016-06-25

摘要/Abstract

摘要：

针对目前Kinect传感器人工标定方法误差大、速度慢等问题，提出一种自动、快速的Kinect传感器外参标定方法。首先，根据彩色图像提取的角点，生成彩色图像的角点集合；其次，为了实现角点点云的自动提取，对点云图像进行深度分割，提取棋盘格点云，采用三维哈夫（Hough）变换检测方法将棋盘格点云投影到深度图像的模板平面上，在深度图像模板中提取深度图像中的角点；然后，将深度图像中的角点映射到棋盘格点云中，形成角点点云；最后，将角点点云与彩色图像的角点集合进行配准，得到角点的3D空间坐标，进而计算出深度相机到彩色相机的姿态变换矩阵。实验结果表明，本文提出的算法在保证相机标定精度的前提下，将相机参数的计算时间从平均218 ms降低到166 ms，实现了自动、快速的Kinect相机标定。

关键词: Kinect, 相机标定, 点云配准, 姿态估计

Abstract:

We propose a quick autocalibration method towards Kinect sensors to solve the problems such as large error and slow computation speed. Firstly, the corner points of the RGB image are extracted to form the corner point set. Then the point cloud of checkboard making from the RGB image is projected using the 3D Hough transform detection method to the depth image in order to autoextract the point cloud corners. Then the point corners in the depth image are extracted. Thirdly, the point cloud corners are projected to the depth image using the least squaring plane fitting algorithm. Then the point cloud corners of the depth images are registered to the corner points of the RGB image. Finally, the 3D coordinates of these corner points are made and the pose transform matrix can be calculated. Experimental results show that the proposed algorithm can extract corner points automatically and decrease the average calibration time from 218 ms to 166ms, thus realizing a quick autocalibration of Kinect sensors.

Key words: Kinect;calibration;point cloud registration;pose estimation

孟勃，刘雪君. 一种自动、快速的Kinect标定方法[J]. J4, 2016, 38(06): 1193-1199.

MENG Bo,LIU Xuejun. A quick autocalibration method of Kinect [J]. J4, 2016, 38(06): 1193-1199.

参考文献［11］

［1］	Smisek J,Jancosek M,Pajdla T.3D with Kinect［M］∥Consumer depth cameras for computer vision.London:Springer, 2013:325.
［2］	Correa D S O,Sciotti D F,Prado M G,et al.Mobile robots navigation in indoor environments using Kinect sensor［C］∥Proc of 2012 IEEE 2rd Brazilian Conference on Critical Embedded Systems, 2012:3641.
［3］	Herrera D,Kannala J,Heikkil J.Accurate and practical calibration of a depth and color camera pair［C］∥Proc of Computer Analysis of Images and Patterns,2011:437445.
［4］	Herrera C,Kannala J,Heikkil J.Joint depth and color camera calibration with distortion correction［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,34(10):20582064.
［5］	Raposo C,Barreto J P,Nunes U.Fast and accurate calibration of a Kinect sensor［C］∥Proc of IEEE International Conference on 3D Vision, 2013:342349.
［6］	Guo Lianpeng, Chen Xiangning, Liu Bin. Calibration of Kinect sensor with depth and color camera［J］.Journal of Image and Graphics,2014,19(11):15841590.(in Chinese)
［7］	Zhang C,Zhang Z.Calibration between depth and color sensors for commodity depth cameras［M］∥Computer vision and machine learning with RGBD sensors.London:Springer International Publishing,2014:4764.
［8］	Liu Xin,Xu Huanrong,Hu Zhanyi.GPU based fast 3Dobject modeling with Kinect［J］.Acta Automatica Sinica,2012,38(8):12881297.(in Chinese)
［9］	Han Yun, Zhong Shenglun, Ye Zhengsheng, et al.Calibration of DRGB camera networks by skeletonbased viewpoint invariance transformation［J］.Acta Physical Sinica,2014,63(7):07421110742119.(in Chinese)
［10］	Hu Zhengchuan.Study on Kinect calibration based RTCam model and space assembling algorithm［D］.Xi’an:Xidian University,2014.(in Chinese)
［11］	Zhang Dayong, Wu Wenqi, Wu Meiping, et al. Plane landmark detection from lidar data based on 3D Hough transform［J］.Journal of National University of Defense Technology,2010,32(2):130134.(in Chinese)
［12］	Gavin H. The LevenbergMarquardt method for nonlinear least squares curvefitting problems［D］.NC,USA:Duke University,2011:115.
	附中文参考文献：
［6］	郭连朋,陈向宁,刘彬.Kinect 传感器的彩色和深度相机标定［J］.中国图象图形学报,2014,19(11):15841590.
［8］	刘鑫,许华荣,胡占义.基于 GPU 和 Kinect 的快速物体重建［J］.自动化学报,2012,38(8):12881297.
［9］	韩云,钟圣伦,叶正圣,等.基于视角无关转换的深度摄像机定位技术［J］.物理学报,2014,63(7):074211:0809.
［10］	胡正川.基于 RTCam 模型的 Kinect 摄像机标定以及空间拼接算法研究［D］.西安：西安电子科技大学,2014.
［11］	章大勇,吴文启,吴美平,等.基于三维 Hough 变换的机载激光雷达平面地标提取［J］.国防科技大学学报,2010,32(2):130134.

[1]	曹玉东, 陈冬昊, 曹睿, 赵朗. 融合Mask R-CNN的在线多目标行人跟踪方法[J]. 计算机工程与科学, 2023, 45(07): 1216-1225.
[2]	周升儒, 陈志刚, 邓伊琴. 基于PoseC3D的网球动作识别及评价方法[J]. 计算机工程与科学, 2023, 45(01): 95-103.
[3]	姚隆兴, 韩江涛, 张志毅. Tsai氏相机平面标定算法的一种解析改进[J]. 计算机工程与科学, 2022, 44(11): 1976-1984.
[4]	黄展原, 李兵, 李庚浩, . 基于视频和人体姿态估计的老年人摔倒监测研究[J]. 计算机工程与科学, 2021, 43(05): 883-890.
[5]	刘国华1,2，段建春1. 基于Kinect相机的深度图像空洞修复及超像素分割算法[J]. 计算机工程与科学, 2020, 42(05): 851-858.
[6]	周游，耿楠，张志毅. 一种改进的稀疏迭代最近点算法[J]. 计算机工程与科学, 2017, 39(10): 1877-1883.
[7]	孟勃1，曹蕾2. 基于Kinect的机器人辅助超声扫描系统研究[J]. J4, 2016, 38(03): 494-500.
[8]	赵建敏,赵忠鑫,许晓伟. 基于Kinect的拳击虚拟训练系统[J]. J4, 2015, 37(09): 1736-1741.
[9]	曹小峰，程实，陈翔，陈晓红. 基于改进Kinect的点分布模型面部表情特征提取[J]. J4, 2015, 37(04): 802-805.
[10]	陈姝1，梁文章2，伍靓1. 基于Kinect深度相机的实时三维人体动画[J]. J4, 2014, 36(08): 1544-1548.
[11]	张艳,赵丽娜,胡万宝. 中心折反射相机标定方法综述[J]. J4, 2014, 36(05): 951-956.
[12]	章权兵，孙庆，王磊，程鸿. 旋转曲面的度量重构[J]. J4, 2013, 35(8): 109-113.
[13]	谭志国1,鲁敏1,胡延平2,郭裕兰1,庄钊文1. 基于点云-模型匹配的激光雷达目标识别[J]. J4, 2012, 34(4): 32-36.
[14]	张根宝,杨永. 粒子滤波器及其在卫星姿态估计中的应用[J]. J4, 2011, 33(6): 159-163.
[15]	汤泽胜,王兆仲. 单帧图像人体姿态估计综述[J]. J4, 2011, 33(11): 89-97.