Large-scale terrain modeling from multiple sensors with dependent Gaussian processes

Terrain modeling remains a challenging yet key component for the deployment of ground robots to the field. The difficulty arrives from the variability of terrain shapes, sparseness of the data, and high degree uncertainty often encountered in large, unstructured environments. This paper presents significant advances to data fusion for stochastic processes modeling spatial data, demonstrated in large-scale terrain modeling tasks. We explore dependent Gaussian processes to provide a multi-resolution representation of space and associated uncertainties, while integrating sensors from different modalities. Experiments performed on multiple multi-modal datasets (3D laser scans and GPS) demonstrate the approach for terrains of about 5 km2.

[1]  Wolfram Burgard,et al.  A Bayesian regression approach to terrain mapping and an application to legged robot locomotion , 2009, J. Field Robotics.

[2]  H. Wackernagle,et al.  Multivariate geostatistics: an introduction with applications , 1998 .

[3]  Marcus R. Frean,et al.  Dependent Gaussian Processes , 2004, NIPS.

[4]  Gamini Dissanayake,et al.  Stochastic simulation in surface reconstruction and application to 3D mapping , 2002, Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292).

[5]  C. Anderson,et al.  Quantitative Methods for Current Environmental Issues , 2005 .

[6]  Ted Chang,et al.  Introduction to Geostatistics: Applications in Hydrogeology , 2001, Technometrics.

[7]  Barak A. Pearlmutter,et al.  Transformations of Gaussian Process Priors , 2004, Deterministic and Statistical Methods in Machine Learning.

[8]  G. Matheron.  Principles of geostatistics , 1963 .

[9]  Ioannis M. Rekleitis,et al.  Experimental Results for Over-the-Horizon Planetary Exploration Using a LIDAR Sensor , 2008, ISER.

[10]  Hugh F. Durrant-Whyte,et al.  Gaussian Process modeling of large scale terrain , 2009, 2009 IEEE International Conference on Robotics and Automation.

[11]  Edwin V. Bonilla,et al.  Multi-task Gaussian Process Prediction , 2007, NIPS.

[12]  Marc Alexa,et al.  Point set surfaces , 2001, Proceedings Visualization, 2001. VIS '01..

[13]  Hugh F. Durrant-Whyte,et al.  Non-stationary dependent Gaussian processes for data fusion in large-scale terrain modeling , 2011, 2011 IEEE International Conference on Robotics and Automation.

[14]  Mark J. Schervish,et al.  Nonstationary Covariance Functions for Gaussian Process Regression , 2003, NIPS.

[15]  Mohammed El-Beltagy,et al.  Gaussian Processes for Model Fusion , 2001, ICANN.

[16]  D. Higdon.  Space and Space-Time Modeling using Process Convolutions , 2002 .

[17]  Simon Lacroix,et al.  Autonomous Rover Navigation on Unknown Terrains: Functions and Integration , 2002, Int. J. Robotics Res..

[18]  Wolfram Burgard,et al.  Multi-Level Surface Maps for Outdoor Terrain Mapping and Loop Closing , 2006, 2006 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[19]  I. Moore,et al.  Digital terrain modelling: A review of hydrological, geomorphological, and biological applications , 1991 .

[20]  Hugh Durrnat-Whyte,et al.  A Critical Review of the State-of-the-Art in Autonomous Land Vehicle Systems and Technology , 2001 .

[21]  Leonidas J. Guibas,et al.  Uncertainty and Variability in Point Cloud Surface Data , 2004, PBG.

[22]  Simon Lacroix,et al.  Autonomous Rover Navigation on Unknown Terrains Functions and Integration , 2000, ISER.

[23]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.