Autonomous Robot Navigation: Path Planning on a Detail-Preserving Reduced-Complexity Representation of 3D Point Clouds

Determination of a collision free, optimal path achieved by performing complex computations on an accurate representation of the terrain, is essential to the success of autonomous navigation systems. This paper defines and builds upon a technique --- Spatially Important Point (SIP) Identification --- for reducing the inherent complexity of range data without discarding any essential information. The SIP representation makes onboard autonomous navigation a viable option, since space and time complexity is greatly reduced. A cost based, dynamic navigation analysis is then performed using only the SIPs which culminates into a new time and space efficient Path Planning technique. The terrain is also retained in the form of a graph, with each branch of every node encountered, indexed in a priority sequence given by its cumulative cost. Experiments show the entire dataflow, from input data to path planning, being done in less than 700ms on a modern computer on datasets typically having 106 points.

[1]  Osamu Takahashi,et al.  Motion planning in a plane using generalized Voronoi diagrams , 1989, IEEE Trans. Robotics Autom..

[2]  Bo Yang,et al.  A Reduction Method of Three-Dimensional Point Cloud , 2009, 2009 2nd International Conference on Biomedical Engineering and Informatics.

[3]  John R. Miller,et al.  Extracting tree crown properties from ground-based scanning laser data , 2007, 2007 IEEE International Geoscience and Remote Sensing Symposium.

[4]  Rodney A. Brooks,et al.  A subdivision algorithm in configuration space for findpath with rotation , 1983, IEEE Transactions on Systems, Man, and Cybernetics.

[5]  Manuela M. Veloso,et al.  SetA*: an efficient BDD-based heuristic search algorithm , 2002, AAAI/IAAI.

[6]  Thierry Siméon,et al.  3-D autonomous navigation in a natural environment , 1994, Proceedings of the 1994 IEEE International Conference on Robotics and Automation.

[7]  N. Dodgson,et al.  A new point cloud simplification algorithm , 2003 .

[8]  Radu Bogdan Rusu,et al.  3D is here: Point Cloud Library (PCL) , 2011, 2011 IEEE International Conference on Robotics and Automation.

[9]  Tim D. Barfoot,et al.  The Canadian planetary emulation terrain 3D mapping dataset , 2013, Int. J. Robotics Res..

[10]  Hongbin Zha,et al.  Scene understanding in a large dynamic environment through a laser-based sensing , 2010, 2010 IEEE International Conference on Robotics and Automation.

[11]  Hongbin Zha,et al.  Computing object-based saliency in urban scenes using laser sensing , 2012, 2012 IEEE International Conference on Robotics and Automation.

[12]  Ashley William Gunter,et al.  Getting it for free: Using Google earth™ and IL WIS to map squatter settlements in Johannesburg , 2009, 2009 IEEE International Geoscience and Remote Sensing Symposium.

[13]  K. Lee,et al.  Data Reduction Methods for Reverse Engineering , 2022 .

[14]  Sven Wachsmuth,et al.  Reducing noise and redundancy in registered range data for planar surface extraction , 2008, 2008 19th International Conference on Pattern Recognition.

[15]  Hans-Peter Kriegel,et al.  A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise , 1996, KDD.