Temporal Matching : Recognizing Dynamic Situations from Discrete Measurements

Abstract

The converse problem of measurement interpretation is event recognition. In situations which are characterized by a specific order of events, a single snapshot is not sufficient to recognize an event. Instead one has to plan a measurement sequence that consists of observations at more than one time point. In this paper we present an algorithm for planning such an observation sequence based on the specification of the event and discuss the problem of giving a meaningful definition of a 'successful match of a measurement sequence against a situation description'.