Use of Haar-like features in vision-based human-computer interaction systems

In everyday life human gestures are often used to communicate or enhance speech. They can be also used to enable or improve the communication between human and machine. Among the contactless human-computer interfaces (HCI) the vision-based solutions enabling for face and hand gesture recognition are the most promising ones. Haar-like object detection algorithm developed by Viola and Jones allows for rapid detection of human faces or hands in image sequences. This paper presents the overview of vision-based Human-Computer interfaces employing methods based on Haar-like features and proposes a Human-Computer Interaction system controlled by mouth shape change.