In today’s information society – when people not only work but also spend a significant part of their free time with computers – there is a growing need for an alternative approach of human-computer interaction (HCI). It has already been shown that the most commonly used input devices (mouse, keyboard) have negative effects on ergonomic and health as well. There is a clearly need of deviceless communication, but the current state of technical issues doesn't allow convenient usage.
The first half of my thesis examines the scope of this problem. After that the features of camera-based approach, related algorithms and procedures are presented. The scenario of hand recognition in 2D images and its dynamic properties are examined due to the input control interpretation.
A framework was prepared and – based on this – a demonstration software was implemented to present the success of the algorithms. My results are shown in the last section.