TinEye: Reverse Image Search Engine

Imagine that you have a poor quality image for which you seek a better quality version to use in your teaching or studies. Perhaps you have explored local databases, ARTstor, Google Images, or other search engines that locate images using keywords, tags, and other metadata, but have not been able to locate the right image. You may want to try a search using TinEye, a freely available reverse image search engine that uses image identification technology to find images on the Web. It allows you to search for images that are visually similar to an image you submit to TinEye. TinEye creates a unique and compact digital signature on your image and retrieves matches by comparing it to all other images in its growing index. At the time of this writing, TinEye has indexed 1,016,014,242 images from the Web. While TinEye can find partial matches, including images that have been resized or edited, it cannot locate similar images based on subject matter.

The Web site lists a few of the many reasons you may wish to use TinEye:

• Find out where an image came from, or get more information about it
• Research or track the appearance of an image online
• Find higher resolution versions of an image
• Locate web pages that make use of an image you have created
• Discover modified or edited versions of an image

You can conduct a search on TinEye's Web site by temporarily uploading an image to the site or by entering a URL of another Web site containing an image of interest. There are also browser plugins for Firefox and Internet Explorer which allow a simple right-click to quickly and easily launch a "Search Image on TinEye" option. TinEye's FAQ page provides a wealth of information for those wanting to know more.

Image: from TinEye's Cool Searches page.