OMITTED

摘要

The present invention provides a sound replaying system based on emotion recognized from facial expressions which comprises: at least one camera for taking images of a subject including at least one person; a facial image output unit for outputting a facial image of the subject from the images output from the camera; an emotion recognition unit for detecting the curvature of the facial image output from the facial image output unit, for producing conformity degrees by comparing the curvature of the facial image with pre-stored patterns such as delight, sad and anger feeling patterns, and for recognizing a calculated emotional pattern, which corresponds to a curvature pattern with the highest degree of conformity, as the present emotional state of the subject; an environmental information processing unit for receiving the information including seasons, weather conditions, temperatures and humidity of a place, on which the subject is located, from an external server; a first emotion judgment unit for judging the emotion of the subject as delight feeling, sad feeling or anger feeling by using a predetermined algorithm based on the current emotional state of the subject which is recognized by the emotion recognition unit and the discomfort index of the place on which the subject is currently located, which is calculated by the environmental information processing unit; a second emotion judgment unit for judging the emotion stored the most as the emotion of at least one subject by making the emotion of the subject, which is judged by the first emotion judgment unit, accumulated and stored therein; a sound storing unit for storing at least one music corresponding to delight, sad and anger feeling; a sound selecting unit for selecting the music corresponding to the emotion judged by the second emotion judgment unit therefrom; and a sound replay unit for playing the music selected by the sound selecting unit.