Hands-Free Gesture and Voice Control for System Interfacing

Abstract

The proposed system presents a simple prototype system for real-time tracking of a human head and speech recognition for hands- free mouse. This system uses a simple yet an effective Face tracking algorithm. The Haar-classifier algorithm is used to capture the frames of the face and Lucas-Kanade algorithm for marking the features of a human face. The general requirements of a real-time tracking algorithm ? it should be computationally economical, should possess the capability to perform in diverse environments and should be able to run itself with a very minimal knowledge about the preexistence of the faces in the head tracking algorithm. This system also makes use of Microsoft Speech SDK 5.1 for speech recognition. It is composed of two fundamental components ? voice recognizer and speech synthesizer. The voice recognizer is used to capture the input of voice signals and speech synthesizer is responsible for lexicon management.