EMOTION RECOGNITION MOTHOD AND SYSTEM BASED ON DECISION FUSION

摘要

A decision fusion based emotion recognition method and a system thereof are provided to perform decision fusion by using a sigmoid fuzzy membership function, thereby complementing a disadvantage of each medium by adding a weight. A decision fusion based emotion recognition method comprises the following steps. A voice signal is collected(P1). First information including at least one of statistics of a pitch, a maximum value, a size of a sound, a number of sections, an increasing rate or a crossing rate is extracted from the voice signal(P2). A face image signal is recognized(P3). Second information including at least one characteristic among a mouth, eyes or eyebrows is extracted(P4). By applying the first and second information to a decision fusion based fuzzy membership function, a pattern by emotion is classified(P5).