An adaptive algorithm for embedded real-time point cloud ground segmentation

This paper presents a fast algorithm for ground segmentation that quickly and accurately differentiates ground points from obstacles after processing unstructured point clouds. Unlike most recent approaches found in the literature, it does not rely on any sensor-specific feature or data ordering. It performs an orthogonal projection into the horizontal plane followed by a top-down 4-ary tree segmentation. The segmentation self-adapts to the point cloud, focusing processing effort on detailed areas. This adaptive subdivision process allows successfully extracting ground points even when the floor is not perfectly flat. Finally, tests demonstrate real-time performance for execution in low cost embedded devices.

[1]  Christoph Stiller,et al.  Segmentation of 3D lidar data in non-flat urban environments using a local convexity criterion , 2009, 2009 IEEE Intelligent Vehicles Symposium.

[2]  M. Himmelsbach,et al.  Real-time object classification in 3D point clouds using point feature histograms , 2009, 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[3]  Philippos Mordohai,et al.  A quantitative evaluation of surface normal estimation in point clouds , 2014, 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[4]  Bertrand Douillard,et al.  On the segmentation of 3D LIDAR point clouds , 2011, 2011 IEEE International Conference on Robotics and Automation.

[5]  Edwin Olson,et al.  Graph-based segmentation for colored 3D laser point clouds , 2010, 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[6]  Armin B. Cremers,et al.  Efficient radius neighbor search in three-dimensional point clouds , 2015, 2015 IEEE International Conference on Robotics and Automation (ICRA).

[7]  Dirk Wollherr,et al.  A clustering method for efficient segmentation of 3D laser data , 2008, 2008 IEEE International Conference on Robotics and Automation.

[8]  Robert C. Bolles,et al.  Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.

[9]  Martial Hebert,et al.  Onboard contextual classification of 3-D point clouds with learned high-order Markov Random Fields , 2009, 2009 IEEE International Conference on Robotics and Automation.

[10]  Beatriz Marcotegui,et al.  Point cloud segmentation towards urban ground modeling , 2009, 2009 Joint Urban Remote Sensing Event.

[11]  Jeffrey Horn,et al.  Handbook of evolutionary computation , 1997 .

[12]  Ji Zhang,et al.  Visual-lidar odometry and mapping: low-drift, robust, and fast , 2015, 2015 IEEE International Conference on Robotics and Automation (ICRA).

[13]  Joshua Hampp,et al.  Quadtree-based polynomial polygon fitting , 2013, 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[14]  A. E. Eiben,et al.  Introduction to Evolutionary Computing , 2003, Natural Computing Series.

[15]  Ben Upcroft,et al.  Advantages of exploiting projection structure for segmenting dense 3D point clouds , 2013, ICRA 2013.

[16]  Reinhard Klein,et al.  Efficient RANSAC for Point‐Cloud Shape Detection , 2007, Comput. Graph. Forum.

[17]  Jaehyun Park,et al.  Quadtree sampling-based superpixels for 3D range data , 2014, 2014 IEEE International Conference on Robotics and Automation (ICRA).

[18]  Mark E. Campbell,et al.  Segmentation of dense range information in complex urban scenes , 2010, 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems.