Dynamic gesture recognition using signal processing based on fuzzy nominal scales
Résumé
During the conversion of the state of a physical entity to an information entity the measurement process preserves a part of the properties of the set of physical states. The set of the preserved properties depends on the initial set of properties and on the scale used for the measurement. It defines the set of relations and operators than can be applied to information entities. This means that the measurement scale induces the set of operators that can be used for signal processing. This paper presents some characteristics of fuzzy nominal scales, more specifically the existence of a distance operator on such scales. Numerical data provided by a data glove are transformed into linguistic data with fuzzy nominal scales. Then the processing of linguistic signals measured with a fuzzy nominal scale is presented. Finally the linguistic data given by the glove are then processed to recognize dynamic hand gestures.