The University of Queensland UQ NavigationUQ HomeUQ SearchUQ MapsUQ ContactsUQ FAQsUQ Library
ITEE Innovation Expo 2002
  World Class: Be Part of It



On this site

  Head of School's Welcome
  UQ Centre Floorplan
  Programme
  Location
  Sponsors
  Student Project List
  Prizes
  2001 Photo Gallery
  Contact Details

Quick Links

  ITEE Innovation Expo 2002
  ITEE Innovation Expo 2001


  Home » Student Projects » Michael Day

Maestro project hardware

Student: Michael Day

Supervisor: Gordon Wyeth

Category: Engineering Thesis Project - Computer Systems

Gesture Based Input Device

With the ever increasing number of electronic devices in our lives new and innovative ways must be found to interact with computers. Ubiquitous computing requires that computers become so much a part of our lives that we rarely become aware of our interactions with them. For this to become a reality however we must be able to converse with these computers in a way that is both un-obtrusive and natural. It is for this reason that this study has been undertaken to develop new ways to interface with computers and other electronic devices.

This prject has further developed a watch and ring design for the recognition of hand gestures. This device allows the control of a television via simple hand gestures and in doing so removes the need to ever look for the remote controll again. Not only can users control a television subconsciously, but they can even control it from another room becuase of the use of radio control over the traditional Infra Red (IR) approach.

The ring is made up of two, two dimensional accelerometers that read in acceleration in 3D. These allow data to be gathered on the acceleration of the finger, so that gestures can be recognised. Once data has been recorded for the gesture it is feed into a Neural Network for processing. Once a gesture has been recognised data is sent over a wireless link to the reciever box where it is then decoded and a command sent to the television.

By using Neural Networks to recognise gestures the device is able to quickly come to a best guess estimate of which gesture was performed and not require a large amount of computing power. This enables the device to fit within the rough dimensions of a watch and ring and not need external computers to process the data.

 

 

feedback
©2001 The University of Queensland, Australia
ABN: 63 942 912 684
Authorised by: Secretary & Registrar
Maintained by: webmasters@itee.uq.edu.au
  Last Updated: 2 July 2001