AUTOMATIC AND SEMI-AUTOMATIC IMAGE CLASSIFICATION, ANNOTATION AND TAGGING THROUGH THE USE OF IMAGE ACQUISITION PARAMETERS AND METADATA

摘要

A kind of characterization image content automatically or semi-automatically uses the presentation of image acquisition parameter and metadata. This method establishes probability and deterministic relationship, in the image of different types of metadata and semantic attribute and content. It provides a kind of mechanism, and permission with semi-automatic classification, annotates, label automatically, and content of the image based on them, the characteristic of the feature of semanteme and metadata are searched for, identify or retrieved to index. This method uses, but is not limited to, image capture metadata such as focal length, time for exposure, relative aperture, flash information, ISO settings, visual angle,, it is used as target range timestamp, GPS information and other metadata, including but not limited to, title, keyword, title, label, comment business remarks, title these can automatically, semi-automatically or manually generate. Present invention could apply to image data bases, Web search, and personal search, community search, extensive or vertical search engine is for internet, Intranet, extranet or other purposes.