Information for topics

Topic Id:

ID topic:	484
Partner Email:	L.J.M.Rothkrantz@tudelft.nl
Project Title:	Screen estimation for a novel pointing device based on corner detection and classification
Abstract:	“UI-Wand” is the name of a new project in the Machine Interface Group of PHILIPS research laboratories in Aachen, Germany. It’s a futuristic concept that uses a pen-shaped pointing device with a camera in its tip to recognize the objects in your home and control them by doing some gestures with it. The first logical step towards its goal is to control applications running on devices with a screen, like Computer, Laptop, or TV. So the basic requirement for current UI-Wand control is the pointing positioning and gesture recognition. This master thesis proposes solutions to several software modules in such a UI-Wand system. Typically, such system includes screen corner detection, screen corners tracking and gesture recognition. Finally we implemented all of the models and algorithms that we chose or designed in the system and implemented a prototype of a UI-Wand driver that demonstrates what can be achieved using such a pointing device.
Advisor:	Leon Rothkrantz
Link:
Degree:	Master
Keywords:
Computer Software
Algorithms & problem solving
Artificial intelligence & Neural networks
Automata & state machines
Computer vision
Data mining
Data modeling
Data types & Data structures
Graphical User Interfaces
Image processing