V-SLAM中点云配准算法改进及移动机器人实验
|
摘要
针对移动机器人视觉同时定位与地图构建(visual simultaneous location and mapping,V-SLAM)中,存在帧间配准误差大造成重建精度低、位姿轨迹丢失的问题,提出一种三阶段改进点云配准的ICP算法.首先通过RANSAC(随机采样一致性)采样策略对RGB图进行点对的筛选从而获得内点完成预处理;然后采用基于刚体变换一致性的对应点双重距离阈值法完成点云初配准;在得到良好的初始位姿下,引入一种动态迭代角度因子的ICP精配准方法.在后端部分引入滑动窗口法和随机采样法相结合的关键帧筛选机制,结合g2o(general graph optimization)图优化算法优化机器人位姿轨迹,实现全局一致的VSLAM系统.采用标准点云模型对本文算法与文献算法进行点云配准实验对比,在配准精度上有明显提高;通过移动机器人在真实环境下的地图重建实验,验证了本文算法的有效性;最后基于TUM数据集的实验表明了本文算法能有效估计出机器人运行轨迹.
A three-stage improved point cloud registration ICP algorithm is proposed to deal with the problem of the low reconstruction accuracy,position trajectory loss caused by matching error in the process of inter-frame registration in visual simultaneous location and mapping( V-SLAM) of mobile robots. Firstly,the RANSAC algorithm is used to eliminate the mis-matching of RGB images in preprocessing phrase. Then,rough registration is accomplished by a double distance threshold method based on rigid body transformation consistency. Under a good initial pose,a dynamic iterative angle factor is used to complete the accurate registration of ICP. On the back end,a key frame screening mechanism combining the sliding window method and the random sampling method is introduced. And the robot pose is optimized by combining the general algorithm optimization( g2 o) algorithm to construct a globally consistent map V-SLAM system. The standard point cloud model is used to verify that the proposed algorithm has a significant improvement in the accuracy of the registration compared experiment. The effectiveness of the proposed algorithm is validated by the map reconstruction experiment in real environment on the mobile robot. Finally,based on the experiment on TUM dataset,the experimental result shows that this algorithm can estimate the robot's trajectory effectively.
A three-stage improved point cloud registration ICP algorithm is proposed to deal with the problem of the low reconstruction accuracy,position trajectory loss caused by matching error in the process of inter-frame registration in visual simultaneous location and mapping( V-SLAM) of mobile robots. Firstly,the RANSAC algorithm is used to eliminate the mis-matching of RGB images in preprocessing phrase. Then,rough registration is accomplished by a double distance threshold method based on rigid body transformation consistency. Under a good initial pose,a dynamic iterative angle factor is used to complete the accurate registration of ICP. On the back end,a key frame screening mechanism combining the sliding window method and the random sampling method is introduced. And the robot pose is optimized by combining the general algorithm optimization( g2 o) algorithm to construct a globally consistent map V-SLAM system. The standard point cloud model is used to verify that the proposed algorithm has a significant improvement in the accuracy of the registration compared experiment. The effectiveness of the proposed algorithm is validated by the map reconstruction experiment in real environment on the mobile robot. Finally,based on the experiment on TUM dataset,the experimental result shows that this algorithm can estimate the robot's trajectory effectively.
引文
[1]ENDRES F,HESS J,STURM J,et al. 3-D mapping with an RGBD camera[J]. IEEE Transactions on Robotics,2014,30(1):177.DOI:10.1109/TRO.2013.2279412
[2]CADENA C,CARLONE L,CARRILLO H,et al. Simultaneous localization and mapping:present,future,and the robust-perception age[J]. IEEE Transaction on Robotics,2016,32(6):1309. DOI:10. 1109/TRO.2016.2624754
[3]FUENTES-PACHECO J,RUIZ-ASCENCIO J,RENDN-MANCHA J M. Visual simultaneous localization and mapping:a survey[J].Artificial Intelligence Review,2015,43(1):55. DOI:10. 1007/s10462-012-9365-8
[4]BESL P J,MCKAY N D. A method for registration of 3-d shapes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1992,14(2):239. DOI:10. 1109/34. 121791
[5]杨扬,曹其新,朱笑笑,等.面向机器人手眼协调抓取的3维建模方法[J].机器人,2013,35(2):151. DOI:10. 3724/SP. J.1218. 2013. 00151YANG Yang,CAO Qixin,ZHU Xiaoxiao,et al. A 3D modeling method for robot’s hand-eye coordinated grasping[J]. Robot,2013,35(2):151. DOI:10. 3724/SP. J.1218.2013.00151
[6]贾松敏,王可,郭兵,等.基于RGB-D相机的移动机器人三维SLAM[J].华中科技大学学报(自然科学版),2014,42(1):103.DOI:10.13245/j.hust.140122JIA Songmin,WANG Ke,GUO Bing,et al. Mobile robot 3D SLAM based on RGB-D camera[J]. Journal of Huazhong University of Science and Technology(Nature Science),2014,42(1):103. DOI:10.13245/j.hust.140122
[7]侯荣波,魏武,黄婷,等.基于ORB-SLAM的室内机器人定位和三维稠密地图构建[J].计算机应用,2017,37(5):1439.DOI:10.11772/j.issn.1001-9081.2017.05.1439HOU Rongbo,WEI Wu,HUANG Ting,et al. Indoor robot localization and 3D dense mapping based on ORB-SLAM[J]. Journal of Computer Applications,2017,37(5):1439. DOI:10. 11772/j. issn.1001-9081.2017.05.1439
[8]LI C,LU B,ZHANG Y,et al. 3D reconstruction of indoor scenes via image registration[J]. Neural Processing Letters,2018(4):1.DOI:10.1007/s11063-018-9781-0
[9]FISCHLER M A,BOLLES R C. Random sample consensus:a paradigm for model fitting with applications to image analysis and automated cartography[J]. Readings in Computer Vision,1987,24(6):726. DOI:10.1145/358669.358692.
[10]ZHANG Z. Microsoft kinect sensor and its effect[J]. IEEE Multimedia,2012,19(2):4. DOI:10. 1109/MMUL. 2012. 24
[11]王忠立,赵杰,蔡鹤皋.大规模环境下基于图优化SLAM的后端优化方法[J].哈尔滨工业大学学报,2015,47(7):75.DOI:10.11918/j.issn.0367-6234.2015.07.002WANG Zhongli,ZHAO Jie,CAI Hegao. A survey of front-end method for graph-based slam under large-scale environment[J].Journal of Harbin Institute of Technology,2015,47(7):75. DOI:10.11918/j.issn.0367-6234.2015.07.002
[12]DANIEL H C,KANNALA J,HEIKKILA J. Joint depth and color camera calibration with distortion correction[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2012,34(10):2058.DOI:10.1109/TPAMI.2012.125
[13]RUBLEE E,RABAUD V,KONOLIGE K,et al. ORB:an efficient alternative to SIFT or SURF[C]//International Conference on Computer Vision.Barcelona:IEEE Computer Society,2011:2564.DOI:10.1109/ICCV.2011.6126544
[14]ROSTEN E,DRUMMOND T. Machine learning for high-speed corner detection[C]//European Conference on Computer Vision. Heidelberg:Springer,2006:430. DOI:10. 1007/11744023_34
[15]CALONDER M,LEPETIT V,STRECHA C,et al. BRIEF:binary robust independent elementary features[C]//European Conference on Computer Vision. Heidelberg:Springer,2010:778. DOI:10.1007/978-3-642-15561-1_56
[16]MUJA M,LOWE D G. Scalable nearest neighbor algorithms for high dimensional data[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2014,36(11):2227. DOI:10. 1109/TPAMI.2014.2321376
[17]STEFAN L,SIMON L,MICHAEL B,et al. Keyframe-based visualinertial odometry using nonlinear optimization[J]. International Journal of Robotics Research,2015,34(3):314. DOI:10.1177/0278364914554813
[18]KARAMAN S,FRAZZOLI E. Sampling-based algorithms for optimal motion planning[J]. Sage Publications,Inc.,2010,30(7):5326. DOI:10. 1177/0278364911406761
[2]CADENA C,CARLONE L,CARRILLO H,et al. Simultaneous localization and mapping:present,future,and the robust-perception age[J]. IEEE Transaction on Robotics,2016,32(6):1309. DOI:10. 1109/TRO.2016.2624754
[3]FUENTES-PACHECO J,RUIZ-ASCENCIO J,RENDN-MANCHA J M. Visual simultaneous localization and mapping:a survey[J].Artificial Intelligence Review,2015,43(1):55. DOI:10. 1007/s10462-012-9365-8
[4]BESL P J,MCKAY N D. A method for registration of 3-d shapes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1992,14(2):239. DOI:10. 1109/34. 121791
[5]杨扬,曹其新,朱笑笑,等.面向机器人手眼协调抓取的3维建模方法[J].机器人,2013,35(2):151. DOI:10. 3724/SP. J.1218. 2013. 00151YANG Yang,CAO Qixin,ZHU Xiaoxiao,et al. A 3D modeling method for robot’s hand-eye coordinated grasping[J]. Robot,2013,35(2):151. DOI:10. 3724/SP. J.1218.2013.00151
[6]贾松敏,王可,郭兵,等.基于RGB-D相机的移动机器人三维SLAM[J].华中科技大学学报(自然科学版),2014,42(1):103.DOI:10.13245/j.hust.140122JIA Songmin,WANG Ke,GUO Bing,et al. Mobile robot 3D SLAM based on RGB-D camera[J]. Journal of Huazhong University of Science and Technology(Nature Science),2014,42(1):103. DOI:10.13245/j.hust.140122
[7]侯荣波,魏武,黄婷,等.基于ORB-SLAM的室内机器人定位和三维稠密地图构建[J].计算机应用,2017,37(5):1439.DOI:10.11772/j.issn.1001-9081.2017.05.1439HOU Rongbo,WEI Wu,HUANG Ting,et al. Indoor robot localization and 3D dense mapping based on ORB-SLAM[J]. Journal of Computer Applications,2017,37(5):1439. DOI:10. 11772/j. issn.1001-9081.2017.05.1439
[8]LI C,LU B,ZHANG Y,et al. 3D reconstruction of indoor scenes via image registration[J]. Neural Processing Letters,2018(4):1.DOI:10.1007/s11063-018-9781-0
[9]FISCHLER M A,BOLLES R C. Random sample consensus:a paradigm for model fitting with applications to image analysis and automated cartography[J]. Readings in Computer Vision,1987,24(6):726. DOI:10.1145/358669.358692.
[10]ZHANG Z. Microsoft kinect sensor and its effect[J]. IEEE Multimedia,2012,19(2):4. DOI:10. 1109/MMUL. 2012. 24
[11]王忠立,赵杰,蔡鹤皋.大规模环境下基于图优化SLAM的后端优化方法[J].哈尔滨工业大学学报,2015,47(7):75.DOI:10.11918/j.issn.0367-6234.2015.07.002WANG Zhongli,ZHAO Jie,CAI Hegao. A survey of front-end method for graph-based slam under large-scale environment[J].Journal of Harbin Institute of Technology,2015,47(7):75. DOI:10.11918/j.issn.0367-6234.2015.07.002
[12]DANIEL H C,KANNALA J,HEIKKILA J. Joint depth and color camera calibration with distortion correction[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2012,34(10):2058.DOI:10.1109/TPAMI.2012.125
[13]RUBLEE E,RABAUD V,KONOLIGE K,et al. ORB:an efficient alternative to SIFT or SURF[C]//International Conference on Computer Vision.Barcelona:IEEE Computer Society,2011:2564.DOI:10.1109/ICCV.2011.6126544
[14]ROSTEN E,DRUMMOND T. Machine learning for high-speed corner detection[C]//European Conference on Computer Vision. Heidelberg:Springer,2006:430. DOI:10. 1007/11744023_34
[15]CALONDER M,LEPETIT V,STRECHA C,et al. BRIEF:binary robust independent elementary features[C]//European Conference on Computer Vision. Heidelberg:Springer,2010:778. DOI:10.1007/978-3-642-15561-1_56
[16]MUJA M,LOWE D G. Scalable nearest neighbor algorithms for high dimensional data[J]. IEEE Transactions on Pattern Analysis&Machine Intelligence,2014,36(11):2227. DOI:10. 1109/TPAMI.2014.2321376
[17]STEFAN L,SIMON L,MICHAEL B,et al. Keyframe-based visualinertial odometry using nonlinear optimization[J]. International Journal of Robotics Research,2015,34(3):314. DOI:10.1177/0278364914554813
[18]KARAMAN S,FRAZZOLI E. Sampling-based algorithms for optimal motion planning[J]. Sage Publications,Inc.,2010,30(7):5326. DOI:10. 1177/0278364911406761
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |