A Low-Pass-Filter-Based Touch-Less Interaction System

摘要

As a recently developed interaction method, touch-less interaction has been utilized in many scenes, from game industry, VR to medical usage and scientific research. Depth-camera-based touch-less interaction, as a new branch of touch-less interaction, is widely welcomed for its low expense and high accuracy. However, current systems are mainly based on specific scenes, with a relatively low precision. We have designed a Kinect-based interaction system for common use with customization options. To improve the accuracy of the captured gesture data, we have designed a feasible low-pass filtering strategy, bringing tremendous enhancement in the accuracy and stability of the system.