This paper presents a technique for real time human hand gesture recognition system for automatic tap operation. Hand gesture recognition based machines have been widely developed in recent years. Most of the techniques use skin color models due to its robustness and simplicity.
In this work there are two major steps for achieving this goal: firstly, the detection of skin color and secondly, the classification of a gesture into valid and invalid hand gesture for the operation of the water tap. The proposed system uses Support Vector Machine and neural network for classification of the hand gesture which in turn operates the tap.