A framework for human-machine interaction using Depthmap and compactness

In recent past research and development in the field of Human-computer/machine interaction (HMI) is on fast track because there has been enough advancement of technologies such as advanced image processing techniques and sensors etc. which provides ease to formulate a test bed to experiment various HMI. A device called Kinect has been introduced for controlling functionalities of windows operating system, which is being widely used in Xbox games by Microsoft. This paper illustrates one of the popular HMI application called mouse control using hand gesture recognition. The method is named as hand-mouse (HM). HM tracks the full body vital core point and the movement of palm controls the movement of mouse cursor and the opening and closing of palm controls a mouse click events. The HM system is using Kinect in order to avoid wearing of any sensors or gesture training dataset, which reduces the computational overhead and experimental results demonstrates a real time output with satisfactory results having higher true positive and lesser false negative.