1. Person Tracking and Pursuit by Mobile Robot - (July - Aug 2006)

   In association with David Rawlinson. Supervisor : Prof. Ray Jarvis, Monash University

    
2. Laser and Panoramic Vision Fusion for Multiple People Tracking - (Oct 2005 - Feb 2006)

   Supervisor : Prof. Ray Jarvis, Monash University

   -Tracking of 3 people walking around a panoramic camera/laser unit.

   -Frame differencing locates targets in panoramic camera and laser images

   -Each target modelled using colour information from panoramic image

   -Each target tracked using a particle filter in laser image.

   -Laser depth information disambiguates occlusions.

    

3. Multiple Target Tracking  -  (May - June 2005)

          Supervisor : Prof. Ray Jarvis, Monash University

          -This project deals with tracking multiple interacting targets over time

          -The number of targets is previously unknown and they do not move under constraints of velocity/acceleration

          -Upto 10 simulated targets have been successfully tracked in real-time, in the presence of heavy clutter

2. Vision-based Indoor GPS for a Motorized Wheelchair -  (Jan - March 2005)

          Project work conducted at CAIR [The Centre for Artificial Intelligence and Robotics, under the

           Defence Research and Development Organization (DRDO)], Bangalore, India

          Supervised by Mr. Sartaj Singh (Scientist "C", CAIR)

          -Achieved indoor localization of a wheelchair on a pre-constructed map, using a camera as the only sensor

          -Used SIFT (Scale Invariant Feature Transform Algorithm) for feature matching between current image

           from the camera and images previously stored in a database.

          -A set of 8 images corresponded to each topological location on the map, and the best matched picture

           indicated the current location of the wheelchair

3. People Tracking for Surveillance Applications -  (March - Nov 2004)

          -Tracking of people over time from a fixed camera using Computer Vision techniques

          -Tracking of mobile robot over time from fixed camera

          -Robot Localization using Particle Filtering of sonar readings

          -Machine-built mapping between the image plane and the ground plane in which the robot moves

          -Robotic intervention to surveillance using fixed camera systems

4. Autonomous Navigation System (March-November,2003)

         Final year project, Dept. Electrical and Electronic Engineering, University of Melbourne

         (completed  Nov 2003)
         Supervised by A/Prof. Mohammed Aldeen (Melbourne EEE)
         -Received mention in an article “Bright Sparks” in the Herald Sun supplement “Connect”, Oct.23,

         2003
         -Finalist for Best Hardware Project at Endeavor 2003, an exhibition showcasing the university’s

          final year electrical and electronic engineering projects

         -Autonomous robot Cyclops built, with Vision capabilities
         -Image Processing Algorithms for recognizing 3-D objects in real-time perfected
         -Navigation Algorithm developed that controls robot, searches for a target whilst avoiding

            obstacles
         -Software developed in Matlab, Java, and C

[ top ]

5. Audio Compression using Fast Fourier Transforms. (March-June, 2002)

         Third Year Project, Dept. Electrical and Electronic Engineering, University of Melbourne 

          (completed June 2002)

        -Supervised by Lecturer Arnaud Doucet (Melbourne EEE; currently at University of Cambridge,

          U.K.)

        -Upto 85% compression of Speech Signals achieved using algorithm developed based on Fast

          Fourier Transforms

[ top ]

6. Robots - Here are a few of the robots that I have built.

                Heimdall  :  This is my current research platform and began life as a Pioneer robot that I have customized.

              Cyclops : A mobile bot capable of navigating itself through an indoor environment and

                                  recognizing objects (part of the Autonomous Navigation System project described

                                  above).

[ top ]

Seminar Presentation -

    1. Data Mining in Images at the First Australian Undergraduate Students' Computing Conference,

        AUSCC 2003. Here are the slides from my presentation at AUSCC.

[ top ]

Online Resources for Robotics

1. Alphabetic List of Robotics and Automation Research Groups

2. Nasa's Mars Rovers

3. Red Team Racing, Carnegie Mellon's entry into the Grand Challenge.

4. Intelligent Robotics Research Centre, Monash Uni.

    

   [ top ]

Resources for Computer Vision

1. Vision pages from the American Association of A.I.

2. IBM Research : People Vision

3. Video tracking Technologies

4. Stanford University class 223b (introduction to computer vision)

5. Visual Surveillance

6. Fundamentals of Computer Vision, University of Maryland

7. Visual Dynamics Research Group, Oxford University

8. Sussex Computer Vision Teach Files

9. People Tracking, University of Reading

10. Finding and tracking people with a hand-held camera, outdoors

11. Tracking people from image sequences, Oulu University, Finland

12. VISOR BASE (VIdeo Sensor Object Request Broker open Architecture for distributed SErvices)

13. Connected Components Analysis

14. CV online: The Evolving, Distributed, Non-Proprietary, On-line Compendium on Computer Vision

15. Image Processing Learning Resources

16. M.I.T Open Courseware in Electrical Engineering and Computer Sciences

17. Looking at People, an overview of research papers

18. Annotated Computer Vision Bibliography: Table of Contents

19. Computer Vision Conference Listing

20. Particle Filters:

1. Sequential Monte Carlo Methods Papers

2. UWashington Particle Filters for Mobile Robotics and People Tracking

3. The Condensation Algorithm

4. Using a Particle filter for Gesture Recognition

5. Multiple object tracking using the Particle Filter

21. Face Detection:

     

    1. Skin probability: Cai, Goshtasby

    2. Face Detection in colour images

    3. Face Detection Techniques

    4. Video Tracking Technologies and Face Detection Systems