Implementation of Human Machine Interface on Raspberry Pi Platform

OData support
Dr. Vajda Ferenc
Department of Control Engineering and Information Technology

The traditional input devices used in computer systems are constantly being replaced by new interaction solutions, that are more natural to use. Some of these require special devices, others try to develop more convenient solutions using only the already available accessories. One widespread example for this is the hand gesture control of smart televisions. Unfortunately, they are rather expensive.

My goal is to develop an interaction system that works like smart televisions, using only one webcam. The most important requirement was to achieve real-time speed in the system. As an extra challenge, I used the Raspberry Pi as the target device, which is a very cheap single-board computer with limited resources, used worldwide.

During my work I researched topics like hand recognition, motion tracking and gesture recognition. After reviewing the existing work in these topics and developing my own algorithms, the result was a hand gesture interaction system that is good enough for daily usage, and can be integrated into more complicated systems, too.


Please sign in to download the files of this thesis.