摘要: This work describes the design of a functioning user interface based on visual recognition hand gestures, and details its performance. In interface, gesture replaces mouse for many actions including selecting, moving resizing windows. A camera below screen observes user. The is segmented from background using color. Features hand's motion are extracted sequence images, when needed pose classified neural net. information parsed by task specific grammar. system runs in real time standard PC hardware. It has demonstrated abilities with various users several different office environments. Having experimented gestural authors discuss practicality best applications this technology.
sci-hub.se 参考文章(6)
LiMin Fu, Neural Networks in Computer Intelligence ,(1994)
David Zeltzer, David Joel Sturman, Whole-hand input Massachusetts Institute of Technology. ,(1992)
Thad E. Starner, Visual Recognition of American Sign Language Using Hidden Markov Models. Massachusetts Institute of Technology. ,(1995)
Jintae Lee, T.L. Kunii, Model-based analysis of hand posture IEEE Computer Graphics and Applications. ,vol. 15, pp. 77- 86 ,(1995) , 10.1109/38.403831
R. Kjeldsen, J. Kender, Finding skin in color images Proceedings of the Second International Conference on Automatic Face and Gesture Recognition. pp. 312- 317 ,(1996) , 10.1109/AFGR.1996.557283
G. Wolberg, Digital Image Warping IEEE Computer Society Press. ,(1990)