Terrain Classification for Ground Robots Based on Acoustic Features

The motivation of our work is to detect different
terrain types traversed by a robot based on acoustic data from the
robot-terrain interaction. Different acoustic features and classifiers
were investigated, such as Mel-frequency cepstral coefficient and
Gamma-tone frequency cepstral coefficient for the feature extraction,
and Gaussian mixture model and Feed forward neural network for the
classification. We analyze the system’s performance by comparing
our proposed techniques with some other features surveyed from
distinct related works. We achieve precision and recall values between
87% and 100% per class, and an average accuracy at 95.2%. We also
study the effect of varying audio chunk size in the application phase
of the models and find only a mild impact on performance.

Authors:

• Bernd Kiefer
• Abraham Gebru Tesfay
• Dietrich Klakow

Keywords:

• Terrain classification
• acoustic features
• autonomous robots
• feature extraction.

References:

[1] Debangshu Sadhukhan, Carl Moore, and Emmanuel Collins, “Terrain
estimation using internal sensors,” in Proceedings of the 10th IASTED
International Conference on Robotics and Applications (RA), 2004.
[2] David Tick, Tauhidur Rahman, Carlos Busso, and Nicholas Gans,
“Indoor robotic terrain classification via angular velocity based
hierarchical classifier selection,” in Robotics and Automation (ICRA),
2012 IEEE International Conference on. IEEE, 2012, pp. 3594–3600.
[3] Eric Coyle, Emmanuel G Collins Jr, and Rodney G Roberts, “Speed
independent terrain classification using singular value decomposition
interpolation,” in Robotics and Automation (ICRA), 2011 IEEE
International Conference on. IEEE, 2011, pp. 4014–4019.
[4] Ayanna Howard and Homayoun Seraji, “Vision-based terrain
characterization and traversability assessment,” Journal of Robotic
Systems, vol. 18, no. 10, pp. 577–587, 2001.
[5] Jann Poppinga, Andreas Birk, and Kaustubh Pathak, “Hough based
terrain classification for realtime detection of drivable ground,” Journal
of Field Robotics, vol. 25, no. 1, pp. 67, 2008. [6] Liang Lu, Camilo Ordonez, Emmanuel G Collins Jr, and Edmond M
DuPont, “Terrain surface classification for autonomous ground vehicles
using a 2d laser stripe-based structured light sensor,” in Intelligent
Robots and Systems, 2009. IROS 2009. IEEE/RSJ International
Conference on. IEEE, 2009, pp. 2174–2181.
[7] Christopher Brooks, Karl Iagnemma, et al., “Vibration-based terrain
classification for planetary exploration rovers,” Robotics, IEEE
Transactions on, vol. 21, no. 6, pp. 1185–1191, 2005.
[8] Christian Weiss, Holger Fr¨ohlich, and Andreas Zell, “Vibration-based
terrain classification using support vector machines,” in Intelligent
Robots and Systems, 2006 IEEE/RSJ International Conference on. IEEE,
2006, pp. 4429–4434.
[9] Chris C Ward and Karl Iagnemma, “Speed-independent vibration-based
terrain classification for passenger vehicles,” Vehicle System Dynamics,
vol. 47, no. 9, pp. 1095–1113, 2009.
[10] Jacqueline Libby and Anthony J. Stentz, “Using sound to classify
vehicle-terrain interactions in outdoor environments,” in Robotics and
Automation (ICRA), 2012 IEEE International Conference on. IEEE,
2012, pp. 3559–3566.
[11] Karl D Iagnemma and Steven Dubowsky, “Terrain estimation for
high-speed rough-terrain autonomous vehicle navigation,” in AeroSense
2002. International Society for Optics and Photonics, 2002, pp. 256–266.
[12] Karl Iagnemma, Hassan Shibly, and Steven Dubowsky, “On-line terrain
parameter estimation for planetary rovers,” in Robotics and Automation,
2002. Proceedings. ICRA’02. IEEE International Conference on. IEEE,
2002, vol. 3, pp. 3142–3147.
[13] Ivana Kruijff-Korbayov´a, Francis Colas, Mario Gianni, Fiora Pirri,
Joachim de Greeff, Koen Hindriks, Mark Neerincx, Petter O¨ gren, Toma´sˇ
Svoboda, and Rainer Worst, “TRADR project: Long-term human-robot
teaming for robot assisted disaster response,” KI-K¨unstliche Intelligenz,
pp. 1–9, 2015.
[14] Theodoros Giannakopoulos, Dimitrios Kosmopoulos, Andreas Aristidou,
and Sergios Theodoridis, “Violence content classification using audio
features,” in Advances in Artificial Intelligence, pp. 502–507. Springer,
2006.
[15] Mark C Wellman, Nino Srour, and David B Hillis, “Feature extraction
and fusion of acoustic and seismic sensors for target identification,” in
AeroSense’97. International Society for Optics and Photonics, 1997, pp.
139–145.
[16] Abraham Gebru Tesfay, “Audio data collection for terrain
classification from sound,” Available for Download under
http://ox6.dfki.de/publications/infostore/10/Bernd%20Kiefer?secret=
694424e304f6dbb14f3b1eaff75b9e83.
[17] Michal Reinstein, Vladimir Kubelka, and Karsten Zimmermann,
“Terrain adaptive odometry for mobile skid-steer robots,” in Robotics
and automation (icra), 2013 ieee international conference on. IEEE,
2013, pp. 4706–4711.
[18] Absolem surveillance and rescue, Available: http://www.bluebotics.com/
mobile-robotics/absolem/ (Accessed 27 March 2017).