Abstract
Recent research suggests intrinsically safe robots, such as through soft limbs or artificial skins, to enable close-quarter human-robot collaboration. Intrinsically safe robots allow for risk-minimized instead of collision-free path planning. Risk-minimized path planning can integrate non-binary knowledge—including obstacle probabilities, robot speed, or data age—into the choice of a robot path. In this contribution, we propose a novel approach to probabilistic obstacle detection on color images that is specifically suited for use in real-time risk-minimized path planning. Our approach enhances an existing neural network for object detection by incorporating spatial coherence via a second neural network and an optimization step inspired by simulated annealing. Finally, a bias towards false-positive obstacle detection allows us to avoid the Sleeping Person Problem for online learning. In our experiments, we show that a GPGPU implementation of our approach can process Full HD images at a soft real-time rate of 15 Hz. We conclude that our probabilistic obstacle detection is fit for use in real-time risk-minimized path planning.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Blackmore L et al (2006) A probabilistic approach to optimal robust path planning with obstacles. In: American control conference
Bouwmans T, El Baf F, Vachon B (2010) Statistical background modeling for foreground detection: a survey. In: Handbook of pattern recognition and computer vision, vol 4. World Scientific Publishing
Broadhurst A, Cipolla R (2000) A statistical consistency check for the space carving algorithm. In: BMVC
Culibrk D et al (2007) Neural network approach to background modeling for video object segmentation. IEEE Trans Neural Netw 18:1614–1627
Culibrk D, Crnojevi V (2010) GPU-based complex-background segmentation using neural networks. In: The Irish machine vision and image processing
Gecks T (2011) Sensorbasierte, echtzeitfähige Online-Bahnplanung für die Mensch-Roboter-Koexistenz, PhD thesis, Universität Bayreuth
Kuhn S (2010) Multi-view reconstruction in-between known environments. Technical report, University of Bayreuth
Lacevic B, Rocco P (2010) Towards a complete safe path planning for robotic manipulators. In: Intelligent robots and systems
Ladikos A, Benhimane S, Navab N (2008) Efficient visual hull computation for real-time 3D reconstruction using CUDA. In: CVPR workshops
Lai A, Yung N (1998) A fast and accurate scoreboard algorithm for estimating stationary backgrounds in an image sequence. In: Symposium on circuits and systems
Laurentini A (1994) The visual hull concept for silhouette-based image understanding. IEEE Trans Pattern Anal Mach Intell 16:150–162
Lenz C et al (2012) Fusing multiple kinects to survey shared human-robot workspaces. Technical Report TUM-I1214, Technische Universität München
Locatelli M (2000) Simulated annealing algorithms for continuous global optimization: convergence conditions. J Optim Theor Appl 104:121–133
Maddalena L et al (2010) A fuzzy spatial coherence-based approach to background foreground separation for moving object detection. Neural Comput 19(2):179–186
Ober-Gecks A, Hänel M, Werner T, Henrich D (2014) Fast multi-camera reconstruction and surveillance with human tracking and optimized camera configurations. In: International symposium on robotics
Salvador J, Casas JR (2008) Shape from probability maps with image-adapted voxelization. In: Workshop on multi-camera and multi-modal sensor fusion
Stengel D, Wiedemann T, Vogel-Heuser B (2012) Efficient 3d voxel reconstruction of human shape within robotic work cells. In: Mechatronics and automation
Sugeno M (1974) Theory of fuzzy integrals and its applications, PhD thesis, Tokyo Institute of Technology
Ulmen J, Cutkosky MR (2010) A robust, low-cost and low-noise artificial skin for human-friendly robots. In: ICRA
Toyama K et al (1999) Wallflower: principles and practice of background maintenance. In: IEEE international conference on computer vision
Werner T, Henrich D (2014) Efficient and precise multi-camera reconstruction. In: International conference on distributed smart cameras
Werner T, Henrich D, Riedelbauch D (2017) Design and evaluation of a multi-agent software architecture for risk-minimized path planning in human-robot workcells. In: Zweiter Kongress Montage Handhabung Industrieroboter
Wurm KM et al (2010) OctoMap: a probabilistic, flexible, and compact 3D map representation for robotic systems. In: Conference on robotics and automation
Acknowledgements
This work has partly been supported by the Deutsche Forschungsgemeinschaft (DFG) under grant agreement He2696/11 SIMERO.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG
About this paper
Cite this paper
Werner, T., Bloeß, J., Henrich, D. (2018). Neural Networks for Real-Time, Probabilistic Obstacle Detection. In: Ferraresi, C., Quaglia, G. (eds) Advances in Service and Industrial Robotics. RAAD 2017. Mechanisms and Machine Science, vol 49. Springer, Cham. https://doi.org/10.1007/978-3-319-61276-8_34
Download citation
DOI: https://doi.org/10.1007/978-3-319-61276-8_34
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-61275-1
Online ISBN: 978-3-319-61276-8
eBook Packages: EngineeringEngineering (R0)