IEEE Robotics & Automation Magazine - December 2013 - 95

Conclusion
In this article, we showed that the proposed visual SLAM algorithm implemented on an embedded system can be successfully applied to resource-limited consumer robots such as
home cleaning robots. The proposed algorithm is made efficient and robust, in particular, by taking advantage of the
orthogonal structure of indoor environments and the invariance property offered by an upward-looking camera.
First, we proposed an algorithmic visual compass that
yields the absolute orientation information of a mobile robot
by utilizing the orthogonal lines of an environment detected
in an image captured from an upward-looking camera. Using
the algorithmic visual compass, we were able to simplify a
SLAM problem to two linear estimation problems (that correct the robot trajectory represented based on a hierarchical
map structure) because the robot orientation error can be significantly reduced. This resulted in the SLAM algorithm to be
operated in real time with the map accuracy well maintained
even in large-scale indoor environments.
Second, in contrast to a frontal view camera, the scale
invariance property of features that goes with the upward-looking camera in most indoor/home environments allowed us to
use only the 2-D image features extracted at the original image
scale. This helped speed up the feature matching process.
Finally, to further reduce the computational burden, we
integrated an additional localization process based on the
standard Kalman filter into the SLAM framework because the
computational complexity of the pose graph optimization
process based on a hierarchical map representation is still
proportional to the number of poses (or nodes) in the graph.
In the proposed method, the pose graph optimization process
was only performed in unoccupied grids. This approach was
possible because feature matching between two adjacent
images captured from an upward-looking camera can be performed regardless of a rotation of the robot about the z axis of
the robot base reference frame.
We hope this article will attract the attention of researchers
in this field toward the advancement of lightweight, low-cost,
yet highly reliable and real-time, embedded SLAMs, promoting a wide spread of SLAM-based consumer robots in the
coming years.
Acknowledgments
This work was supported in part by the Future IT Laboratory
of LG Electronics Inc., Korea, KORUS-Tech Program
(KT-2008-SW-AP-FSO-0004) of the Ministry of Knowledge
Economy, Korea, and NRF-2013M1A3A02042335, the
Ministry of Science ICT and Future Planning, Korea.
References
[1] H. Bay, A. Ess, T. Tuytelaars, and L. Van Gool, "Speeded-up robust features,"
Int. J. Comput. Vis. Image Understand., vol. 110, no. 3, pp. 346-359, 2008.
[2] J. Nocedal and S. Wright, Numerical Optimization. New York: SpringerVerlag, 1999.
[3] S. Lee, S. Lee, and J. J. Yoon, "Illumination-invariant localization based on
upward looking scenes for low-cost indoor robots," Adv. Robot., vol. 26, no. 13,
pp. 1443-1469, 2012.

[4] A. J. Davison, "Real-time simultaneous localisation and mapping with a
single camera," in Proc. IEEE Int. Conf. Computer Vision, Beijing, China, 2003,
pp. 1403-1410.
[5] L. A. Clemente, A. J. Davison, I. D. Reid, J. Neira, and J. D. Tardós,
"Mapping large loops with a single hand-held camera," in Proc. Robotics:
Science Systems, Atlanta, GA, 2007.
[6] S. Thrun, M. Bennewitz, W. Burgard, A. Cremers, F. Dellaert, D. Fox, D.
Hähnel, C. Rosenberg, N. Roy, J. Schulte, and D. Schulz, "MINERVA: A second generation mobile tour-guide robot," in Proc. IEEE Int. Conf. Robotics
Automation, Detroit, MI, 1999, pp. 1999-2005.
[7] H. F. Durrant-Whyte and T. Bailey, "Simultaneous localization and mapping: Part I," IEEE Robot. Automat. Mag., vol. 13, no. 2, pp. 99-110, 2006.
[8] T. Bailey and H. F. Durrant-Whyte, "Simultaneous localization and mapping: Part II," IEEE Robot. Automat. Mag., vol. 13, no. 2, pp. 108-117, 2006.
[9] S. Thrun, W. Burgard, and D. Fox, Probabilistic Robotics. Cambridge, MA:
MIT Press, 2005.
[10] R. Hartley and A. Zisserman, Multiple View Geometry in Computer
Vision. Cambridge, MA: Cambridge Univ. Press, 2000.
[11] J. Folkesson, P. Jensfelt, and H. Christensen, "Vision SLAM in the measurement subspace," in Proc. IEEE Int. Conf. Robotics Automation, Barcelona,
Spain, 2005, pp. 30-35.
[12] J. Montiel and A. J. Davison, "A visual compass based on SLAM," in Proc.
IEEE Int. Conf. Robotics Automation, Orlando, FL, 2006, pp. 1917-1922.
[13] G. Grisetti, C. Stachniss, S. Grzonka, and W. Burgard, "A tree parameterization for efficiently computing maximum likelihood maps using gradient
descent," in Proc. Robotics: Science Systems, Atlanta, GA, 2007.
[14] Z. Zhang, "Determining the epipolar geometry and its uncertainty: A
review," Int. J. Comput. Vis., vol. 27, no. 2, pp. 161-195, 1998.
[15] C. Harris and M. Stephens, "A combined corner and edge detector," in
Proc. 4th Alvey Vision Conf., Manchester, U.K., 1988, pp. 147-151.
[16] F. Lu and E. Milios, "Globally consistent range scan alignment for environment mapping," Auton. Robot., vol. 4, no. 4, pp. 333-349, 1997.
[17] W. Y. Jeong and K. M. Lee, "CV-SLAM: A new ceiling vision-based SLAM
technique," in Proc. IEEE/RSJ Int. Conf. Intelligent Robots Systems, Edmonton,
AB, Canada, 2005, pp. 3195-3200.
[18] J. M. M. Montiel, J. Civera, and A. J. Davison, "Unified inverse depth
parametrization for monocular SLAM," in Proc. Robotics: Science Systems,
Cambridge, CA, 2006, pp. 16-19.
[19] N. Karlsson, E. D. Bernardo, J. Ostrowski, L. Goncalves, P. Pirjanian, and
M. E. Munich, "The vSLAM algorithm for robust localization and mapping," in
Proc. IEEE Int. Conf. Robotics Automation, Barcelona, Spain, 2005, pp. 24-29.
[20] B. Steder, G. Grisetti, C. Stachniss, and W. Burgard, "Visual SLAM for
flying vehicles," IEEE Trans. Robot., vol. 24, no. 5, pp. 1088-1093, 2008.
[21] P. Rybski, F. Zacharias, J. Lett, O. Masoud, M. Gini, and N. Papanikolopoulos,
"Using visual features to build topological maps of indoor environments," in Proc.
IEEE Int. Conf. Robotics Automation, Taipei, Taiwan, 2003, pp. 850-855.
[22] A. Koenig, J. Kessler, and H.-M. Gross, "A graph matching technique for an
appearance-based, visual SLAM-approach using Rao-Blackwellized particle filters,"
in Proc. IEEE Int. Conf. Intelligent Robots Systems, Nice, France, 2008, pp. 1576-1581.
[23] K. Konolige, "SLAM via variable reduction from constraint maps," in Proc.
IEEE Int. Conf. Robotics Automation, Barcelona, Spain, 2005, pp. 667-672.
[24] K. Konolige and M. Agrawal, "FrameSLAM: From bundle adjustment to realtime visual mapping," IEEE Trans. Robot., vol. 24, no. 5, pp. 1066-1077, 2008.
[25] S. Lee, E. Kim, and Y. Park, "3D object recognition using multiple features
for robotic manipulation," in Proc. IEEE Int. Conf. Robotics Automation,
Orlando, FL, 2006, pp. 3768-3774.

Seongsoo Lee, Future IT Laboratory, LG Electronics Inc., Seoul,
Korea. E-mail: seongsoo007@gmail.com.
Sukhan Lee, School of Information and Communication
Engineering and Department of Interaction Science,
Sungkyunkwan University, Suwon, Korea. E-mail: lsh@ece.
skku.ac.kr.
DECEMBER 2013

*

IEEE ROBOTICS & AUTOMATION MAGAZINE

*

95
http://skku.ac.kr
Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - December 2013
