This study focuses on a novel touchless interface for mobile devices by using infrared LED sensors. The reflected light from a set of proximity sensors are used to determine the position of the hand through the trilateration method. The position accuracy is easily affected by undesired object. In this paper, the problem can be overcome by using the Kalman filter; the Kalman filter to improve the infrared-based touchless positioning system with dynamic adjustment of measurement noise features. The design offers an intuitive way to specify the commands with mobile devices in real time. At a fixed reference point, the positioning standard deviations of the x-axis, y-axis and z-axis are 0.03 cm, 0.02 cm and 0.01 cm, respectively.