Paper Title
Text Based Image Search Methods A Survey
Abstract
The invention of cheaper digital cameras, computer systems, Internet and fast growth of digital technology results
in digitization of all kinds of materials. The volume of digital images produced by scientific, educational, medical,
industrial, and other applications available to users increased dramatically. Visual information plays major role in
understanding surroundings. Images are powerful tools in communication. Every day, people upload tremendous amount of
images on the web. Mostly image retrieval systems use some method of adding metadata such as keywords / tags, captions
or descriptions to the images so that image retrieval can be performed over the annotation words, called Tag/Text based
information retrieval. TBIR help users to retrieve images efficiently and effectively. But, since users are not serious about
choosing appropriate tags, or assigning tags to the images, create inconsistency, ambiguity in image database which degrades
the performance of TBIR system. Now, large scale image search is a big challenge for image search engines like Google,
Flickr.
Keywords- Image Retrieval Systems, Metadata, TBIR.