摘要 |
A method and a device for generating metadata of contents are provided to reflect emotion information on meta information of the contents by automatically obtaining the emotion information of the contents from a user appreciating the contents. An image input unit(110) obtains a user image whenever each scene of contents is played. A face image extractor(122) extracts face image from the obtained user image. A face image normalizer(124) fits the extracted face image to a size of facial expression template information(132) stored in a storage unit(130). An emotion information generator(126) generates emotion information(134) for the contents by comparing the normalized face image with the facial expression template information. A transmitter transmits the generated emotion information to a server providing the contents. The emotion information generator generates the emotion information including a type and strength of the emotion, which is obtained when the user image is obtained, by referring to the facial expression template information. |