Computer interfaces using vision based gesture recognition
MetadataShow full item record
AbstractThe primary objective of this thesis is to obtain information from a non-instrumented hand - including precise points of contact between the real hand and the virtual environment, and the recognition of meaningful hand postures - so that they can be used to replace standard input devices. Two different methods are presented in this thesis, ERSolitaire and finger classification, which recognize hand postures for use in a natural interface. ERSolitaire provides a grab-move-release interface to the standard game of solitaire, which allows the user to interact with virtual cards in a similar fashion as they would with real cards. The finger classification system provides a virtual touch interface by individually recognizing each of the five fingers. This provides greater power and flexibility than other similar systems, as well as a new technique for hand posture recognition.
Bibliography: p. 138-150