Department of Computer Science and Electrical Engineering
Abstract:
As computing power and imaging technology have become much more accessible
in recent years, it is now possible to bring tangible benefits from image
processing to the desktop - to change the way you work with your computer.
Using a single video camera positioned near your display, MIME (Mime Is
Manual Expression) allows you to use intuitive hand gestures to control
your computer and escape the tangle of mice and keyboards.
The MIME system is a process for recognizing hand gestures captured
using a video camera and a standard consumer personal computer. Initially
developed and implemented using the MATLAB mathematical environment, the
system is designed for implementation in C code for the Microsoft Windows
environment allowing real-time gesture control of new and existing applications.
Using a novel two-dimensional model of the human hand, MIME employs
a efficient, single-pass algorithm to segment a hand and extract its model
parameters from each frame in the video input. The hand is tracked from
one frame to the next using a constant-acceleration Kalman filter approach.
Tracking and feature extraction is robust and efficient.
Complete thesis:
thesis.pdf
Conference Paper:
Conference Paper.pdf