Twins: A practical vision-based 3D mouse

A novel scheme is introduced that uses a mirror and a single camera to restore the 3D position of a finger tip. The camera is positioned in such a way that it captures both the hand as well as its mirror image. The captured images are then processed to extract the contour of the hand. This extraction process is done using a fast algorithm that integrates Bayesian estimation in the traditional chain code generation algorithm. The algorithm works well with a complex, albeit fixed, background, In addition, we have integrated the smoothness assumption into the algorithm so as to obtain a smoother hand contour, thus improving the accuracy of our tracking process. Tolerance to noise can be improved by utilizing the constraints imposed by our system setting to eliminate false signals. A prototype system has been implemented and the performance of the 3D mouse before different backgrounds has been analysed. (C) 1998 Academic Press.