Self-organizing Computer Vision for Robust Object Tracking in Smart Cameras

Computer vision is one of the key research topics of modern computer science and finds application in manufacturing, surveillance, automotive, robotics, and sophisticated human-machine-interfaces. These applications require small and efficient solutions which are commonly provided as embedded systems. This means that there exist resource constraints, but also the need for increasing adaptivity and robustness. This paper proposes an autonomic computing framework for robust object tracking. A probabilistic tracking algorithm is combined with the use of multi-filter fusion of redundant image filters. The system can react on unpredictable changes in the environment through self-adaptation. Due to resource constraints, the number of filters actively used for tracking is limited. By means of self-organization, the system structure is re-organized to activate filters adequate for the current context. The proposed framework is designed for, but not limited to, embedded computer vision. Experimental evaluations demonstrate the benefit of the approach.

[1]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[2]  Michael Isard,et al.  CONDENSATION—Conditional Density Propagation for Visual Tracking , 1998, International Journal of Computer Vision.

[3]  J. March.  Exploration and exploitation in organizational learning , 1991, STUDI ORGANIZZATIVI.

[4]  Walter Stechele,et al.  Hardware/software architecture of an algorithm for vision-based real-time vehicle detection in dark environments , 2008, 2008 Design, Automation and Test in Europe.

[5]  Bernt Schiele,et al.  Towards Robust Multi-cue Integration for Visual Tracking , 2001, ICVS.

[6]  Jürgen Teich,et al.  Self-organizing multi-cue fusion for FPGA-based embedded imaging , 2009, 2009 International Conference on Field Programmable Logic and Applications.

[7]  Zoran A. Salcic,et al.  Customizing Multiprocessor Implementation of an Automated Video Surveillance System , 2006, EURASIP J. Embed. Syst..

[8]  Jason Schlessman,et al.  Hardware/Software Co-Design of an FPGA-based Embedded Tracking System , 2006, 2006 Conference on Computer Vision and Pattern Recognition Workshop (CVPRW'06).

[9]  Wayne H. Wolf,et al.  Smart Cameras as Embedded Systems , 2002, Computer.

[10]  Hartmut Schmeck,et al.  Towards a quantitative notion of self-organisation , 2007, 2007 IEEE Congress on Evolutionary Computation.

[11]  Anthony Rowe,et al.  CMUcam3: An Open Programmable Embedded Vision Sensor , 2007 .

[12]  Bernhard Rinner,et al.  Real-time video analysis on an embedded smart camera for traffic surveillance , 2004, Proceedings. RTAS 2004. 10th IEEE Real-Time and Embedded Technology and Applications Symposium, 2004..

[13]  Anton van den Hengel,et al.  Probabilistic Multiple Cue Integration for Particle Filter Based Tracking , 2003, DICTA.

[14]  Jürgen Teich,et al.  ReCoBus-Builder — A novel tool and technique to build statically and dynamically reconfigurable systems for FPGAS , 2008, 2008 International Conference on Field Programmable Logic and Applications.

[15]  Jochen Triesch,et al.  Democratic Integration: Self-Organized Integration of Adaptive Cues , 2001, Neural Computation.

[16]  Wolfgang Straßer,et al.  Adaptive Probabilistic Tracking Embedded in Smart Cameras for Distributed Surveillance in a 3D Model , 2007, EURASIP J. Embed. Syst..

[17]  Matthias Werner,et al.  On the Definitions of Self-Managing and Self-Organizing Systems , 2011 .

[18]  Vladimir Vezhnevets,et al.  A Survey on Pixel-Based Skin Color Detection Techniques , 2003 .