Twins: A practical vision-based 3D mouse

A novel scheme is introduced that uses a mirror and a single camera to rest ore 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 i mages are then processed to extract the contour of the hand. This extractio n process is done using a fast algorithm that integrates Bayesian estimatio n in the traditional chain code generation algorithm. The algorithm works w ell with a complex, albeit fixed, background, In addition, we have integrat ed the smoothness assumption into the algorithm so as to obtain a smoother hand contour, thus improving the accuracy of our tracking process. Toleranc e to noise can be improved by utilizing the constraints imposed by our syst em setting to eliminate false signals. A prototype system has been implemen ted and the performance of the 3D mouse before different backgrounds has be en analysed. (C) 1998 Academic Press.