"Digitale Bibliothek", Digitale Horizonte

Schriftgröße: 
Information discovery using search by image on the Internet: a brief overview
Paul Nieuwenhuysen

Zuletzt geändert: 2019-03-08

Abstract


‘Search by image’ or ‘reverse image searching’ on the Internet / WWW is a relatively new method, in which a search query consists not of text, but of an image file. The search results lead to related images and further also to related documents on the Internet / WWW. This method of searching can be applied to cope with several types of information needs for which more classical search methods fail or perform less efficiently. TinEye was a pioneering system dedicated to search by image, which can still be used. More recently the more versatile and popular company Google has also offered “reverse image search”; this performs relatively well. A comparison of visual search as offered by Google and Pinterest has recently shown that Pinterest has become an additional and competing system that allows us to start from a selected source image to find related, relevant images. The Google search system even allows a query that includes both text and an image; this can yield results with a higher precision than more simple queries. Applications include finding more suitable versions of your source image, revealing re-usage of your images, detecting misleading information items and retrieving information that is related to the contents of your source image.

 

 

PRÄSENTATION:

Information discovery using search by image on the Internet: a brief overview. (.pdf)

Paul Nieuwenhuysen