Real-Time Map Building with Uncertainty using Colour Camera and Scanning Laser

This paper presents a real-time data fusion technique to produce three dimensional colour point clouds of the environment from a mobile platform. Range from a scanning laser and colour from a video camera are locally fused using deterministic transforms. Platform pose information is used to provide global coordinates. Measurement uncertainty is maintained throughout the transformations to enable the fusion of multiple data sets obtained at different times or from different locations, and to enable applications that use the data to assess its quality. The complete system is implemented on an autonomous outdoor ground vehicle and vast colour models of outdoor terrain are built in real time. Although similar techniques have been described in the literature, they have not been applied to a large scale outdoor environment from a mobile platform. The contribution of this paper is to show that very high quality three dimensional colour point clouds can be built in real-time, with a relatively simple approach.

[1]  Robert Pless,et al.  Extrinsic calibration of a camera and laser range finder (improves camera calibration) , 2004, 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (IEEE Cat. No.04CH37566).

[2]  Paul Newman,et al.  Motion Estimation from Map Quality with Millimeter Wave Radar , 2006, 2006 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[3]  Eric Nettleton,et al.  Decentralised Architectures for tracking and navigation with multiple flight vehicles , 2003 .

[4]  Hugh F. Durrant-Whyte,et al.  A Model for Machine Perception in Natural Environments , 2003, ISRR.

[5]  Sebastian Thrun Winning the DARPA grand challenge , 2006 .

[6]  Automated texture mapping of 3D city models with oblique aerial imagery , 2004, Proceedings. 2nd International Symposium on 3D Data Processing, Visualization and Transmission, 2004. 3DPVT 2004..

[7]  Paul J. Besl,et al.  A Method for Registration of 3-D Shapes , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  Kazunori Ohno,et al.  Dense 3D map building based on LRF data and color image fusion , 2005, 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[9]  Mark E. Campbell,et al.  A mixture‐model based algorithm for real‐time terrain estimation , 2006, J. Field Robotics.

[10]  Zoller Fröhlich REALISTIC 3D RECONSTRUCTION – COMBINING LASERSCAN DATA WITH RGB COLOR INFORMATION , 2004 .

[11]  Christopher Rasmussen,et al.  Combining laser range, color, and texture cues for autonomous road following , 2002, Proceedings 2002 IEEE International Conference on Robotics and Automation (Cat. No.02CH37292).

[12]  L. Chrisman,et al.  Obstacle Avoidance and Safeguarding for a Lunar Rover , 1996 .

[13]  THE UNIVERSITY OF SYDNEY , 1957 .

[14]  Steve Scheding,et al.  Calibration of range sensor pose on mobile platforms , 2007, 2007 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[15]  Jean-Yves Bouguet,et al.  Camera calibration toolbox for matlab , 2001 .

[16]  Wolfram Burgard,et al.  An Efficient Extension of Elevation Maps for Outdoor Terrain Mapping , 2005, FSR.