| 摘要 |
<p>Keyframes extracted from a video or blobs (binary large objects) extracted from a multimedia or hypermedia document, are stored so as to optimize retrieval from a relatively slow memory device. Retrieval and storage may be based on significant or representative images in the video or on user preferences on information, including text, audio, video and hyperlinks to other documents, from a multimedia document. Elimination of redundant information is also possible.</p> |
