Expert system for gesture recognition in terminal's user interface

This paper presents several AI (Artificial Intelligent) based embedded gesture recognition procedures for a user interface of a mobile terminal. In the presented solutions, a terminal includes three acceleration sensors positioned like xyz co-ordinate system in order to get three-dimensional acceleration vector, xyz (gesture). The acceleration vector is used as an input to a reasoning unit for classification. Compared classification methods are fuzzy rule aided, fuzzy c-means, SOM (Self Organising Map) and hidden Markov model methods. The developed fuzzy rule aided method classified most successfully test sets. Other advantages of it, are computational effectiveness, lower data sample rate requirement, simple implementation and high reliability. Therefore, the gesture recognition can be applied to the terminals to replace traditional keyboard functions. The computational effectiveness and low sample rate requirement also increases the operational time of device. The easy implementation and reliability are important factors for the success of the new technology's spreading on the mass market of terminals.