Real-time skeleton tracking for embedded systems
作者： Foti Coleca , Sascha Klement , Thomas Martinetz , Erhardt Barth
DOI: 10.1117/12.2003004
关键词:
摘要: Touch-free gesture technology is beginning to become more popular with consumers and may have a signicant future impact on interfaces for digital photography. However, almost every commercial software framework pose detection aimed at either desktop PCs or high-powered GPUs, making mobile implementations recognition an attractive area research development. In this paper we present algorithm hand skeleton tracking that runs ARM-based platform (Pandaboard ES, OMAP 4460 architecture). The uses self-organizing maps t given topology (skeleton) into 3D point cloud. This novel way of approaching the problem as it does not employ complex optimization techniques data-based learning. After initial background segmentation step, ran in parallel heuristics, which detect correct artifacts arising from insucient erroneous input data. We then optimize ARM using xed-point computation NEON SIMD architecture OMAP4460 provides. tested two dierent depth-sensing devices (Microsoft Kinect, PMD Camboard). For both were able accurately track native framerate cameras.
harvard.edu
uni-luebeck.de
sci-hub.se 参考文章(12)
Marc M. Sebrechts, Kay M. Stanney, Deborah M. Clawson, George K. Hung, Marilyn Mantei Tremaine, Corinna E. Lathan, Joseph Wilder, Robert S. Kennedy, Manpreet Kaur, Kristyne E. Kennedy, David Gross, Kathleen M. Bartlett, Michael R. Tracey, Jerry Isdale, Clive Fencott, Michael Heim, HANDBOOK OF VIRTUAL ENVIRONMENTS ,(2009)
Torsten Bierz, Jorg Meyer, Achim Ebert, GPU Accelerated Gesture Detection for Real Time Interaction Visualization of Large and Unstructured Data Sets. pp. 64- 75 ,(2007)
Alex Couture-Beil, Greg Mori, Mark Bayazit, Real-time Motion-based Gesture Recognition Using the GPU. Journal of Machine Vision and Applications. pp. 9- 12 ,(2009)
Mathias Kölsch, Vladimir Pavlović, Branislav Kisačanin, Thomas S. Huang, Special issue on vision for human–computer interaction Computer Vision and Image Understanding. ,vol. 108, pp. 1- 3 ,(2007) , 10.1016/J.CVIU.2007.01.002
Martin Haker, Martin Böhme, Thomas Martinetz, Erhardt Barth, Self-Organizing Maps for Pose Estimation with a Time-of-Flight Camera Lecture Notes in Computer Science. ,vol. 5742, pp. 142- 153 ,(2009) , 10.1007/978-3-642-03778-8_11
A. Kolb, E. Barth, R. Koch, R. Larsen, Time-of-Flight Cameras in Computer Graphics Computer Graphics Forum. ,vol. 29, pp. 141- 159 ,(2010) , 10.1111/J.1467-8659.2009.01583.X
Jamie Shotton, Toby Sharp, Alex Kipman, Andrew Fitzgibbon, Mark Finocchio, Andrew Blake, Mat Cook, Richard Moore, Real-time human pose recognition in parts from single depth images Communications of The ACM. ,vol. 56, pp. 116- 124 ,(2013) , 10.1145/2398356.2398381
W.T. Freeman, K. Tanaka, J. Ohta, K. Kyuma, Computer vision for computer games international conference on automatic face and gesture recognition. pp. 100- 105 ,(1996) , 10.1109/AFGR.1996.557250
Nobutaka Shimada, Kousuke Kimura, Yoshiaki Shirai, None, Real-time 3D hand posture estimation based on 2D appearance retrieval using monocular camera international conference on computer vision. pp. 23- 30 ,(2001) , 10.1109/RATFG.2001.938906
Ali Erol, George Bebis, Mircea Nicolescu, Richard D. Boyle, Xander Twombly, Vision-based hand pose estimation: A review Computer Vision and Image Understanding. ,vol. 108, pp. 52- 73 ,(2007) , 10.1016/J.CVIU.2006.10.012