In June of 2008, I wrote about a new image search engine called TinEye that was claiming to do true image recognition search using pictures as queries.

At the time, the site was claiming:

“TinEye is the first image search engine on the web to use image identification technology. Given an image to search for, TinEye tells you where and how that image appears all over the web—even if it has been modified.

Just as you are familiar with entering text in a regular search engine such as Google to find web pages that contain that text, TinEye lets you submit an image to find web pages that contain that image.”

And, just as promised… the search engine is now publicly available and fully functional.  And, to at least my partial surprise, it works fairly flawlessly.  Here’s an example:

http://tineye.com/search/fd852df5478eb7eb9410ee9101bb364adf487fb0?page=1360

I searched the Google logo straight from their homepage using the image URL as my query.  By the end of the results, it becomes clear that Tineye recognizes the content of the Google logo and not just the color/shape attributes of the letters— there are some images where the logo is skewed/distorted in a way that TinEye must truly understand some of the fundamental aspects of the logo itself.

So, I give TinEye (and its parent company Idée) a big thumbs up for progressing things.  Now, if someone would just figure out how to monetize image search in this way.  That itself may prove to be more difficult than producing a working image search engine.