The developing embedded technology requires revolutions in human-machine interaction. In this paper, we propose a novel method using localization of the taping sound on the table to replace the keyboard as manual input device. The method is applicable with a quad-channel-array collection of acoustic signals, from which the time-of-arrival differences and the position information could be estimated. In practice, as our table is in a limited size and the material properties are complex, the traditional localization algorithm based on time-of-arrival differences contains a sizable margin for error. Furthermore, we use neural-network analysis to improve recognition accuracy. Then experiments and simulations are carried out to verify this signal processing algorithm.

Date Published: 14 March 2013
PDF: 5 pages
Proc. SPIE 8768, International Conference on Graphic and Image Processing (ICGIP 2012), 87686V (14 March 2013); doi: 10.1117/12.2008191

Published in SPIE Proceedings Vol. 8768:
International Conference on Graphic and Image Processing (ICGIP 2012)
Zeng Zhu, Editor(s)