A novel 6-degree of freedom man-machine interface using MEMS sensors and computer vsion

Developing user friendly Human-Computer-Interfaces is important in many areas of computer applications. Traditional input devices such as the keyboard and mouse are designed for general purpose computers only, they are not likely to provide intuitive user experiences for multi-dimensional input entry. Therefore, we design and implement an embedded system for users to interactive with a computer in 6-degree of freedom. The system can capture the 3-D rotation and 3-D translation motions which are the basic 6-D motion parameters. In our proposed approach, the translational motion is estimated using a computer vision method by tracking the speckled pattern generated by a laser beam on a rough surface. The orientation information is estimated by measuring the gravitational and magnetic fields of the Earth. A gyroscope is included together with the use of Kalman filter to improve the dynamic responses. A prototype is built to evaluate the performance of the proposed tracking method. Experiments have been carried out which show that the approach can work with good accuracy and the whole system can be built using off-the-shelf low cost components.