Suppose you have an image and you want to find similar images on the web? This kind of search is known as image content search, or reverse image search. The technology to do this has been gradually maturing over the years, and at last it seems to be working usefully.
Image search engine TinEye is open for business (although still in beta, as at March 2009). You can either upload an image from your computer, or you can enter the URL of any image on the web. TinEye then searches its claimed index of over a billion images, and returns the images that match.
TinEye is not fooled by images that have been cropped, had their colors changed, had writing added, had elements swapped out, or been distorted. It’s impressive to set the search order to “worst match first” to see some truly remarkable matches. You can also set the search order to “best match first”, “largest image first” or “smallest image first”.
A nifty “compare images” link brings up a panel where you can switch between your image and a matching image. The website adjusts the scale and position so that the matched areas overlap exactly, and you can animate between them.
Here’s an example.
Need research? Quezi's researchers can answer your questions at uclue.com