Swarm robotics
PRESENTATION OVERVIEW
 Introduction To Gestures
 Design Of The Proposed System
 Hand Detection
 Hardware Implementation
 Conclusion / Future Work
Real World
Insect
Examples
Natural swarms
 Decentralised – no-one in control
 Individuals are simple and autonomous
 Local communication and control
 Cooperative behaviours emerge through self-
organisation
e.g. repairing damage to nest, foraging for food,
caring for brood
Bees
An In-depth Look at Real
Ant Behaviour
Interrupt The Flow
The Path Thickens!
Adapting to Environment Changes
The New Shortest Path
Adapting to Environment Changes
Welcome to the
Real World
Robots
• Collective task completion
• No need for overly complex
algorithms
• Adaptable to changing environment
Swarm robotics
 Inspired by self-organisation of social insects
 Using local methods of control and communication
Local control: autonomous operation
Local communication: avoids bottlenecks
Scalable – new robots can be added, or fail without need for
recalibration
Simplicity – cheap, expendable robots
 Self-organisation
 Decentralisation
Collective Robotics
Swarm Robotics
Introduction To Gestures
 Gestures can originate from any body
motion.
 Commonly from face/hand.
 Gesture recognition-understand human
body language.
 Help human to interact with machines
without any mechanical devices.
Disadvantages of centralised
control and communication.
 Central control: failure of controller implies
failure of whole system
 Robot to robot communication becomes very
complex as number of robots increases.
 Communication bottlenecks
 Adding new robots means changing the
communication and control system
Design Of The Proposed System
 Hand detection
 Feature extraction
 Gesture recognition
 Goal directed navigation of swarm
robots.
Hand Detection
 Hand detection
– Detection of hand in an image, background objects are
avoided for feature extraction.
– Skin color is the key component.
– Detecting skin and non-skin.
– Detecting image pixels and regions that contains skin-
tone color.
– Background is controlled.
– Appearance depends on illumination conditions.
 Two phases
– Training phase
– Detection phase
Training Phase
 Three steps
– Collecting a database of skin patches from different
images
– Choosing a suitable color space
– Learning the parameters of skin classifier
Detection Phase
 Two steps
Converting the image into some color space that was
used in training phase.
Classifying each pixel using the skin classifier to either
a skin or non-skin.
 RGB color space
 Skin classifier
Variety of classification techniques
Any pixel which color falls inside the skin color class
boundary is labeled as skin.
Feature Extraction
 Feature-An interesting part of an image.
 No exact definition.
 Depends on the problem.
 Transforming the input data into set of features.
 Result is a feature vector.
 Features extracted are invariant to image scaling,
rotation and less affected to changes in
illumination.
 SIFT feature extraction.
 Hand tracking and Feature extraction.
Image frame
from webcam
Tracking
hand by
skin detection
SIFT feature
extraction
Finding
match
Gesture
database
Gesture1 Gesture2 Gesture3
Action1 Action2 Action3
Navigation of
swarm
Hardware Implementation
 Foot-bot robot
Applications Of The System
 Three foot-bots.
Applications of swarm approach
Some tasks are particularly suited to group of expendable
simple robots
e.g. - cleaning up toxic waste
- exploring an unknown planet
- pushing large objects
- surveillance and other military applications
Conclusion
 Hand detection and feature extraction removes
noise from the image.
 System performance and accuracy will
increase.
 Swarm robots movement can be controlled
through gestures.
Future Work
Speech recognition.
Dumb parts, properly
connected into a swarm,
yield smart results.
Kevin Kelly
Satellite
Maintenance
The Future?
Medical
Interacting
Chips in
Mundane Objects
Cleaning Ship
Hulls
Pipe
Inspection
Pest Eradication
M
iniaturization
EngineMaintenance
Telecommunications
Self-Assem
bling
Robots
Job Scheduling
Vehicle
Routing
Data
Clustering
Distributed
M
ail
System
s
O
ptim
al
Resource
Allocation
Combinatorial
Optimization
1. C.C.Wang, K.C.Wang.: Hand Posture Recognition Using Adaboost with SIFT For Human
Robot Interaction, in Robotics: Viable Robotic Service to Human, Springer-2009.
2. M. Kolsch and M. Turk.: Robust hand detection, in IEEE International Conference on
Automatic Face and Gesture Recognition, 2004.
3. Cristina Manresa, Javier Varona, Ramon Mas and Francisco J.Perales,.: Hand Tracking
and Gesture Recognition for Human-Computer Interaction , Electronic Letters on
Computer Vision and Image Analysis 5(3):96-104, 2005.
4. Alessandro Giusti, Jawad Nagi, Luca M. Gambardella, Gianni A. Di Caro : Distributed
Consensus for Interaction between Humans and Mobile Robot Swarms (Demonstration).
5. Ihab Zaqout, Roziati Zainuddin, Sapian Baba,.: Pixel-Based Skin Color Detection
Technique, in Machine Graphics and Vision, 2005. FLEXChip Signal Processor
(MC68175/D), Motorola, 1996.
6. Qiu-yu Zhang, Mo-yi Zhang, Jian-qiang Hu,.: Hand Gesture Contour Tracking Based on
Skin Color Probability and State Estimation Model, Journal of Multimedia, Vol. 4, No. 6,
December 2009. A. Karnik, “Performance of TCP congestion control with rate feedback:
TCP/ABR and rate adaptive TCP/IP,” M. Eng. thesis, Indian Institute of Science,
Bangalore, India, Jan. 1999.
7. Lars Bretzner, Ivan Laptev, Tony Lindberg,.: Hand Gesture Recognition using Multi-Scale
Colour Features, Hierarchical Models and Particle Filtering , Proceedings of the Fifth
IEEE International Conference on Automatic Face and Gesture Recognition (2002).
Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specification,
IEEE Std. 802.11, 1997.
References
Contd…
• G. Lee and N. Y. Chong, “Decentralized formation control for small scale
robot teams with anonymity,” Mechatronics, vol. 19, no. 1, pp.85–
105,2009.
• G. Lee and N. Y. Chong, “A geometric approach to deploying robot
swarms,” Ann. Math. Artif. Intell., vol. 52, no. 2–4, pp. 257–280, 2009.
• E. Sahin, “Swarm robotics: From sources of inspiration to domains of
• application,” in Proc. 8th Int. Conf. Simulation of Adaptive
Behavior(LNCS),2005,vol.3342,pp.10–20.
• H. Niwa, K. Kodaka, Y. Sakamoto, M. Otake, S. Kawaguchi, K. Fujii, Y.
Kanemori, and S. Sugano, “GPS-based indoor positioning system with
multi-channel pseudolite,” in Proc. IEEE Int. Conf. Robot. Autom., 2008,
pp. 905–910. Faraj Alhwarin, Chao Wang, Danijela Risti -Durrant, Axel
Graser, Improved SIFT-Features Matching for Object Recognition, BCS
International Academic Conference- Visions of Computer Science,2008.
Thank you

Swarm ROBOTICS

  • 1.
  • 2.
    PRESENTATION OVERVIEW  IntroductionTo Gestures  Design Of The Proposed System  Hand Detection  Hardware Implementation  Conclusion / Future Work
  • 3.
  • 4.
    Natural swarms  Decentralised– no-one in control  Individuals are simple and autonomous  Local communication and control  Cooperative behaviours emerge through self- organisation e.g. repairing damage to nest, foraging for food, caring for brood
  • 5.
  • 6.
    An In-depth Lookat Real Ant Behaviour Interrupt The Flow
  • 7.
    The Path Thickens! Adaptingto Environment Changes
  • 8.
    The New ShortestPath Adapting to Environment Changes
  • 9.
  • 10.
    Robots • Collective taskcompletion • No need for overly complex algorithms • Adaptable to changing environment
  • 11.
    Swarm robotics  Inspiredby self-organisation of social insects  Using local methods of control and communication Local control: autonomous operation Local communication: avoids bottlenecks Scalable – new robots can be added, or fail without need for recalibration Simplicity – cheap, expendable robots  Self-organisation  Decentralisation
  • 12.
  • 13.
    Introduction To Gestures Gestures can originate from any body motion.  Commonly from face/hand.  Gesture recognition-understand human body language.  Help human to interact with machines without any mechanical devices.
  • 14.
    Disadvantages of centralised controland communication.  Central control: failure of controller implies failure of whole system  Robot to robot communication becomes very complex as number of robots increases.  Communication bottlenecks  Adding new robots means changing the communication and control system
  • 15.
    Design Of TheProposed System  Hand detection  Feature extraction  Gesture recognition  Goal directed navigation of swarm robots.
  • 16.
    Hand Detection  Handdetection – Detection of hand in an image, background objects are avoided for feature extraction. – Skin color is the key component. – Detecting skin and non-skin. – Detecting image pixels and regions that contains skin- tone color. – Background is controlled. – Appearance depends on illumination conditions.  Two phases – Training phase – Detection phase
  • 17.
    Training Phase  Threesteps – Collecting a database of skin patches from different images – Choosing a suitable color space – Learning the parameters of skin classifier
  • 18.
    Detection Phase  Twosteps Converting the image into some color space that was used in training phase. Classifying each pixel using the skin classifier to either a skin or non-skin.  RGB color space  Skin classifier Variety of classification techniques Any pixel which color falls inside the skin color class boundary is labeled as skin.
  • 19.
    Feature Extraction  Feature-Aninteresting part of an image.  No exact definition.  Depends on the problem.  Transforming the input data into set of features.  Result is a feature vector.  Features extracted are invariant to image scaling, rotation and less affected to changes in illumination.  SIFT feature extraction.
  • 20.
     Hand trackingand Feature extraction.
  • 21.
    Image frame from webcam Tracking handby skin detection SIFT feature extraction Finding match Gesture database Gesture1 Gesture2 Gesture3 Action1 Action2 Action3 Navigation of swarm
  • 22.
  • 24.
    Applications Of TheSystem  Three foot-bots.
  • 25.
    Applications of swarmapproach Some tasks are particularly suited to group of expendable simple robots e.g. - cleaning up toxic waste - exploring an unknown planet - pushing large objects - surveillance and other military applications
  • 26.
    Conclusion  Hand detectionand feature extraction removes noise from the image.  System performance and accuracy will increase.  Swarm robots movement can be controlled through gestures. Future Work Speech recognition.
  • 27.
    Dumb parts, properly connectedinto a swarm, yield smart results. Kevin Kelly
  • 28.
    Satellite Maintenance The Future? Medical Interacting Chips in MundaneObjects Cleaning Ship Hulls Pipe Inspection Pest Eradication M iniaturization EngineMaintenance Telecommunications Self-Assem bling Robots Job Scheduling Vehicle Routing Data Clustering Distributed M ail System s O ptim al Resource Allocation Combinatorial Optimization
  • 29.
    1. C.C.Wang, K.C.Wang.:Hand Posture Recognition Using Adaboost with SIFT For Human Robot Interaction, in Robotics: Viable Robotic Service to Human, Springer-2009. 2. M. Kolsch and M. Turk.: Robust hand detection, in IEEE International Conference on Automatic Face and Gesture Recognition, 2004. 3. Cristina Manresa, Javier Varona, Ramon Mas and Francisco J.Perales,.: Hand Tracking and Gesture Recognition for Human-Computer Interaction , Electronic Letters on Computer Vision and Image Analysis 5(3):96-104, 2005. 4. Alessandro Giusti, Jawad Nagi, Luca M. Gambardella, Gianni A. Di Caro : Distributed Consensus for Interaction between Humans and Mobile Robot Swarms (Demonstration). 5. Ihab Zaqout, Roziati Zainuddin, Sapian Baba,.: Pixel-Based Skin Color Detection Technique, in Machine Graphics and Vision, 2005. FLEXChip Signal Processor (MC68175/D), Motorola, 1996. 6. Qiu-yu Zhang, Mo-yi Zhang, Jian-qiang Hu,.: Hand Gesture Contour Tracking Based on Skin Color Probability and State Estimation Model, Journal of Multimedia, Vol. 4, No. 6, December 2009. A. Karnik, “Performance of TCP congestion control with rate feedback: TCP/ABR and rate adaptive TCP/IP,” M. Eng. thesis, Indian Institute of Science, Bangalore, India, Jan. 1999. 7. Lars Bretzner, Ivan Laptev, Tony Lindberg,.: Hand Gesture Recognition using Multi-Scale Colour Features, Hierarchical Models and Particle Filtering , Proceedings of the Fifth IEEE International Conference on Automatic Face and Gesture Recognition (2002). Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specification, IEEE Std. 802.11, 1997. References
  • 30.
    Contd… • G. Leeand N. Y. Chong, “Decentralized formation control for small scale robot teams with anonymity,” Mechatronics, vol. 19, no. 1, pp.85– 105,2009. • G. Lee and N. Y. Chong, “A geometric approach to deploying robot swarms,” Ann. Math. Artif. Intell., vol. 52, no. 2–4, pp. 257–280, 2009. • E. Sahin, “Swarm robotics: From sources of inspiration to domains of • application,” in Proc. 8th Int. Conf. Simulation of Adaptive Behavior(LNCS),2005,vol.3342,pp.10–20. • H. Niwa, K. Kodaka, Y. Sakamoto, M. Otake, S. Kawaguchi, K. Fujii, Y. Kanemori, and S. Sugano, “GPS-based indoor positioning system with multi-channel pseudolite,” in Proc. IEEE Int. Conf. Robot. Autom., 2008, pp. 905–910. Faraj Alhwarin, Chao Wang, Danijela Risti -Durrant, Axel Graser, Improved SIFT-Features Matching for Object Recognition, BCS International Academic Conference- Visions of Computer Science,2008.
  • 31.

Editor's Notes

  • #7 We want to talk in depth about how the algorithm works here