Abstract: The Leap Motion sensor can detect the position and speed information of hands and fingers with high precision in real-time, which provides an effective way for volume interaction with large screen at long-distance. A set of understandable gesture for volume interaction is designed by observing gestures of the users to control the remote real objects. By dividing the near and far field regions of the Leap Motion sensor, a reasonable mapping between physics space and information space is established. According to the estimation solution for hand orientation with Depth Camera, an orientation estimation algorithm combining hand normal vector with fingertip vector is proposed using Leap Motion sensor. In addition, a non touching volume interaction prototype is designed for large screen application, and the prototype evaluation is given through the user experiment. Experimental results show that the proposed algorithm can provide a more natural, intuitive and efficient user experience compared with the traditional 2D volume interaction.