[1]王新杰,张莹,张东波,等.自主移动机器人路径规划中的点云噪声处理[J].智能系统学报,2021,16(4):699-706.[doi:10.11992/tis.202007040]
WANG Xinjie,ZHANG Ying,ZHANG Dongbo,et al.Point cloud noise processing in path planning of autonomous mobile robot[J].CAAI Transactions on Intelligent Systems,2021,16(4):699-706.[doi:10.11992/tis.202007040]
点击复制
WANG Xinjie,ZHANG Ying,ZHANG Dongbo,et al.Point cloud noise processing in path planning of autonomous mobile robot[J].CAAI Transactions on Intelligent Systems,2021,16(4):699-706.[doi:10.11992/tis.202007040]
自主移动机器人路径规划中的点云噪声处理
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
16
期数:
2021年第4期
页码:
699-706
栏目:
学术论文—智能系统
出版日期:
2021-07-05
- Title:
- Point cloud noise processing in path planning of autonomous mobile robot
- Keywords:
- point cloud map; pass through filtering; voxel filtering; statistical filtering; point cloud segmentation; mobile robot; motion planning; RGB-D perception
- 分类号:
- TP242
- 摘要:
- 机器人在自主导航过程中，图像处理算法通过对路径中障碍物的三维点云平面夹角进行分析，实时计算障碍物的高度，判断是否可以通行。而RGB-D相机输出的原始点云数据存在大量噪声，严重影响分割的准确性，需要进行滤波处理。本文在Table Scene数据集中分析比较了统计滤波中不同参数的去噪效果，得到了最佳参数K=20，α=2。移动机器人通过ORB-SLAM2算法在户外环境下构建点云地图，然后分别进行直通滤波、体素滤波、统计滤波和平面分割，计算斜坡夹角，实施运动规划。实验结果表明，在Table Scene数据集中获得的最优统计滤波参数能适用于户外环境，机器人能根据运算结果自动进行路径规划，完成指定任务。
- Abstract:
- In the process of robots’ autonomous navigation, the image processing algorithm analyzes the 3D point cloud plane angles of obstacles in the path and calculates their height in real time, thus determining whether the robots can pass or not. However, a large amount of noise is present in the original point cloud data output by the RGB-D camera, thereby seriously affecting the accuracy of segmentation; such noise thus needs to be filtered. This paper first obtains the best parameters (K=20 and α=2) by analyzing and comparing the denoising effects of different parameters in statistical filtering in the Table Scene dataset. The mobile robot constructs a point cloud map in an outdoor environment through the ORB-SLAM2 algorithm and then performs passthrough filtering, voxel filtering, statistical filtering, and plane segmentation, calculating the slope angle and implementing motion planning. Experimental results show that the optimal statistical filter parameters obtained in the Table Scene dataset can be applied to outdoor environments and that robots can automatically plan their paths according to the calculation results and complete their specified tasks.
- 参考文献/References:
-
[1] 叶一飞, 王建中. 基于点云的复杂环境下楼梯区域识别[J]. 电子测量与仪器学报, 2020, 34(4): 124-133
YE Yifei, WANG Jianzhong. Stair area identification in complex environment based on point cloud[J]. Journal of electronic measurement and instrumentation, 2020, 34(4): 124-133
[2] 李茂月, 马康盛, 王飞, 等. 基于结构光在机测量的叶片点云预处理方法研究[J]. 仪器仪表学报, 2020, 41(8): 55-66
LI Maoyue, MA Kangsheng, WANG Fei, et al. Research on the preprocessing method of blade point cloud based on structured light on-machine measurement[J]. Chinese journal of scientific instrument, 2020, 41(8): 55-66
[3] CHEN Siyuan, TRUONG-HONG L, O’KEEFFE E, et al. Outlier detection of point clouds generating from low-cost UAVs for bridge inspection[C]//Proceedings of the 6th International Symposium on Life-Cycle Civil Engineering. Ghent, Belgium, 2018.
[4] ALEXA M, BEHR J, COHEN-OR D, et al. Computing and rendering point set surfaces[J]. IEEE transactions on visualization and computer graphics, 2003, 9(1): 3-15.
[5] CHEN Shenen, RICE C, BOYLE C, et al. Small-format aerial photography for highway-bridge monitoring[J]. Journal of performance of constructed facilities, 2011, 25(2): 105-112.
[6] ZHANG Yingjie, GE Liling. Improved moving least squares algorithm for directed projecting onto point clouds[J]. Measurement, 2011, 44(10): 2008-2019.
[7] BERGER M, TAGLIASACCHI A, SEVERSKY L, et al. State of the art in surface reconstruction from point clouds[C]//Eurographics 2014-State of the Art Reports. Strasbourg, France, 2014: 161-185.
[8] NURUNNABI A, WEST G, BELTON D. Outlier detection and robust normal-curvature estimation in mobile laser scanning 3D point cloud data[J]. Pattern recognition, 2015, 48(4): 1404-1419.
[9] ZAMAN F, WONG Y P, NG B Y. Density-based denoising of point cloud[M]//IBRAHIM H, IQBAL S, TEOH S S, et al. 9th International Conference on Robotic, Vision, Signal Processing and Power Applications. Singapore: Springer, 2017: 287-295.
[10] MUR-ARTAL R, TARDóS J D. ORB-SLAM2: an open-source SLAM system for monocular, stereo, and RGB-D cameras[J]. IEEE transactions on robotics, 2017, 33(5): 1255-1262.
[11] MUR-ARTAL R, MONTIEL J M M, TARDóS J D. ORB-SLAM: a versatile and accurate monocular SLAM system[J]. IEEE transactions on robotics, 2015, 31(5): 1147-1163.
[12] WOLFF K, KIM C, ZIMMER H, et al. Point cloud noise and outlier removal for image-based 3D reconstruction[C]//Proceedings of 2016 Fourth International Conference on 3D Vision. Stanford, CA, USA, 2016: 118-127.
[13] 祝继华, 周颐, 王晓春, 等. 基于图像配准的栅格地图拼接方法[J]. 自动化学报, 2015, 41(2): 285-294
ZHU Jihua, ZHOU Yi, WANG Xiaochun, et al. Grid Map merging approach based on image registration[J]. Acta automatica sinica, 2015, 41(2): 285-294
[14] 王任栋, 徐友春, 齐尧, 等. 一种鲁棒的城市复杂动态场景点云配准方法[J]. 机器人, 2018, 40(3): 257-265
WANG Rendong, XU Youchun, QI Yao, et al. A robust point cloud registration method in urban dynamic environment[J]. Robot, 2018, 40(3): 257-265
[15] LIU Dingning, DING Qiong. Point cloud processing system development based on PCL and Qt[J]. International journal of engineering management, 2019, 3(1): 33-39.
[16] TRUONG-HONG L, LAEFER D F. Quantitative evaluation strategies for urban 3D model generation from remote sensing data[J]. Computers & graphics, 2015, 49: 82-91.
备注/Memo
收稿日期:2020-07-24。
基金项目:国家自然科学基金项目(61175075)；国家自然科学基金区域创新发展联合基金项目(U19A2083)；湖南省战略性新兴产业科技攻关与重大成果转化项目(2019GK4007)
作者简介:王新杰，硕士研究生，主要研究方向为模式识别与图像处理、视觉导航;张莹，副教授，湖南省自动化学会理事，主要研究方向为机器人控制、视觉导航。主持和参与国家自然科学基金项目、湖南省科技厅项目、湖南省教育厅项目等多项。发表学术论文30余篇;张东波，教授，主要研究方向为模式识别与图像处理、深度学习。湖南省自动化学会副理事长、湖南省青年骨干教师。近5年，承担和参与国家级、省部级课题10余项。发表学术论文100余篇，获授权发明专利7项
通讯作者:张莹.E-mail:zhangying@xtu.edu.cn
更新日期/Last Update:
1900-01-01