An Improved MbICP Algorithm for Mobile Robot Pose Estimation

This paper presents an improved version of the metric-based iterative closest point algorithm to estimate robot poses by matching 2D laser scans with different overlapping percentages. Because of the greatly varied density distribution of realistic point clouds, a resampling method is used to accelerate the iteration process and protect the calculation of the rejection threshold from being distorted by reducing dense but unhelpful points. A new procedure that combines point-to-point and point-to-line metrics is used to determine the correct correspondence between partially overlapping scans, which maintains both efficiency and accuracy. In addition, a rejection threshold that is based on the MAD-from-median method is utilized to discard correspondences with large distances, which are likely to be incorrect. Experiments show that the improved algorithm is more accurate and robust than the standard algorithm with respect to the existence of non-overlapping areas, and testing demonstrates that it is valid in practice.

[1]  Stergios I. Roumeliotis,et al.  Weighted range sensor matching algorithms for mobile robot displacement estimation , 2002, Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292).

[2]  P. R. McAree,et al.  How do ICP variants perform when used for scan matching terrain point clouds? , 2017, Robotics Auton. Syst..

[3]  Antoni Burguera,et al.  Robust scan matching localization using ultrasonic range finders , 2005, 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[4]  François Michaud,et al.  Relative Motion Threshold for Rejection in ICP Registration , 2009, FSR.

[5]  Gérard G. Medioni,et al.  Object modeling by registration of multiple range images , 1991, Proceedings. 1991 IEEE International Conference on Robotics and Automation.

[6]  Evangelos E. Milios,et al.  Robot Pose Estimation in Unknown Environments by Matching 2D Range Scans , 1997, J. Intell. Robotic Syst..

[7]  Heinrich Niemann,et al.  A refined ICP algorithm for robust 3-D correspondence estimation , 2003, Proceedings 2003 International Conference on Image Processing (Cat. No.03CH37429).

[8]  Nanning Zheng,et al.  Point-to-line metric based Iterative Closest Point with bounded scale , 2009, 2009 4th IEEE Conference on Industrial Electronics and Applications.

[9]  Ran Liu,et al.  Outlier Robust ICP for Minimizing Fractional RMSD , 2006, Sixth International Conference on 3-D Digital Imaging and Modeling (3DIM 2007).

[10]  Roland Siegwart,et al.  Comparing ICP variants on real-world data sets , 2013, Auton. Robots.

[11]  Evangelos E. Milios,et al.  Robot Pose Estimation in Unknown Environments by Matching 2D Range Scans , 1994, 1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[12]  S. Druon,et al.  Color Constrained ICP for Registration of Large Unstructured 3D Color Data Sets , 2006, 2006 IEEE International Conference on Information Acquisition.

[13]  Soon-Yyong Park,et al.  Global localization for mobile robots using reference scan matching , 2014 .

[14]  Basilio Bona,et al.  Scan Matching for Graph SLAM in Indoor Dynamic Scenarios , 2014, FLAIRS Conference.

[15]  Hugh F. Durrant-Whyte,et al.  A solution to the simultaneous localization and map building (SLAM) problem , 2001, IEEE Trans. Robotics Autom..

[16]  Karsten Berns,et al.  Simultaneous localization and mapping (SLAM) , 2009 .

[17]  Lei Zhang,et al.  Robust ICP Registration Using Biunique Correspondence , 2011, 2011 International Conference on 3D Imaging, Modeling, Processing, Visualization and Transmission.

[18]  Donghwan Kim,et al.  Spectral scan matching for robot pose estimation , 2009 .

[19]  K. S. Arun,et al.  Least-Squares Fitting of Two 3-D Point Sets , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[20]  Deyu Meng,et al.  Robust registration of partially overlapping point sets via genetic algorithm with growth operator , 2014, IET Image Process..

[21]  Roland Siegwart,et al.  Scan alignment with probabilistic distance metric , 2004, 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (IEEE Cat. No.04CH37566).

[22]  Luis Montesano,et al.  Probabilistic scan matching for motion estimation in unstructured environments , 2005, 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[23]  Yisha Liu,et al.  Mobile robot instant indoor map building and localization using 2D laser scanning data , 2012, 2012 International Conference on System Science and Engineering (ICSSE).

[24]  Takanori Emaru,et al.  A solution to estimate robot motion with large rotation by matching laser scans , 2015, 2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan (SICE).

[25]  Pavel Krsek,et al.  The Trimmed Iterative Closest Point algorithm , 2002, Object recognition supported by user interaction for service robots.

[26]  Florent Lamiraux,et al.  Metric-based iterative closest point scan matching for sensor displacement estimation , 2006, IEEE Transactions on Robotics.

[27]  Lindsay Kleeman,et al.  Fast Laser Scan Matching using Polar Coordinates , 2007, Int. J. Robotics Res..

[28]  Xiaoyong Wang,et al.  Mobile robot pose estimation using laser scan matching based on Fourier Transform , 2013, 2013 IEEE International Conference on Robotics and Biomimetics (ROBIO).

[29]  Bum-Jae You,et al.  Robust laser scan matching in dynamic environments , 2009, 2009 IEEE International Conference on Robotics and Biomimetics (ROBIO).

[30]  Marc Levoy,et al.  Efficient variants of the ICP algorithm , 2001, Proceedings Third International Conference on 3-D Digital Imaging and Modeling.

[31]  Nuno Lau,et al.  Efficient localization based on scan matching with a continuous likelihood field , 2017, 2017 IEEE International Conference on Autonomous Robot Systems and Competitions (ICARSC).