| 摘要 |
<P>PROBLEM TO BE SOLVED: To speedily extract desired data with inexpensive configuration. <P>SOLUTION: A base level detection unit 60 is configured to accept input of image and audio data for one frame in input data and to detect a feature amount of images or audio for one frame, a low level detection unit 70 detects information of whether a scene change occurs in contents or not, or the like as intermediate data on the basis of base level output data output from the base level detection unit 60, and a high level data detection unit 80 detects information for specifying phases of highlight scene, CM and the like in the contents, for example, as metadata on the basis of low level output data output from the low level detection unit 70. <P>COPYRIGHT: (C)2009,JPO&INPIT |
