High Level Cognitive Vision for Developing Smart Environment using Hand Gesture Recognition

Abstract

As technology is growing tremendously electronic
gadgets are becoming social agents in our day today life. Nowadaysconcept of Smart environment where human can interact withmachine using its biological intelligence is developing. Biologicalintelligence means interacting machine with his natural way of communication using speech, gestures etc.. Moreover Human–computer/machine interaction (HCI) systems capable of sensing and responding to the user‟s natural way of interaction is likely to be perceived as more natural , more efficacious and persuasive . With
the help of this paper a mode of communication which is away from data entry by keyboard, mouse etc. , using hand gestures is tried to develop. The current state of art is that hardware interfaces like gloves etc. are being designed but to form rich vocabulary and grammar for operating machine which will give a wireless and natural way of interaction with machine is still a challenging task. In this paper high level classification of complicated hand gesture behavior, by which human wants to issue a command is simplified by combining statistical and structural pattern recognition approaches. These hand gesture behavior which are converted into binary using image processing tools, forms words of the vocabulary, which are then interpreted as various control commands for various devices. Using combination of two approaches shows that vocabulary of command instruction can be increased easily, simultaneously using image features to develop context free grammar helps to solve the problem of forming gestural language.