Published On: Sat, Sep 14th, 2013

Visual Search Is Moving Beyond Google

You know when you have an image saved to your computer or posted to your Facebook wall and, for the life of you, cannot remember where you got it? Or when you see a screencapture from a movie that looks pretty cool, but nobody knows which movie it’s from? Reverse image search can be a life saver in these instances, helping you to find the source of an image, or to find similar images. These image search engines can deliver some surprising results if you take the time to play around with them.

Female hand reaching images streaming


Retrievr is just about the most fun you’re going to have with a search engine today. Go to the site and you’ll find a simple Microsoft Paint-style sketch pad. You can doodle whatever you like, and seconds later have dozens of matching images pop up. It is, however, difficult to get precise results. You can’t draw a caricature of Brad Pitt and get a bunch of photos of the actor (unless you’re a truly incredible artist), but the engine does a fine job of finding images based on color and basic composition. Scribble some blue and white and you’ll get images of breezy summer skies. Scribble red on black and get some great campfires at night. At this point, the engine is probably more useful for art projects and as a fun diversion, but search-by-sketch offers an exciting glimpse of a more naturalistic, organic tool than the current word-game feel of Google and Bing.


There are a lot of different approaches to building an image search engine. As notes, while Google Image search is good for finding images by keyword and reverse image searches, it was not built for images. Where Retrievr is more of a toy, TinEye is more of a tool. If you’re a blogger, if you’re in publishing, or if you need to rely on images for what you do, then you need to bookmark TinEye. It’s a great resource for anyone who relies on images. If you have a great photo that you want to put in a blog post, TinEye can help you to find a higher resolution version of the same picture, if it exists. It can tell you where the image came from. It can tell you how other people are using the image so that you don’t wind up treading the same ground. It even serves as the backbone of editorial image monitoring service PixID, and an essential tool in eBay auctions.


Chromatik is perfect for web designers and graphic artists. You can search by any combination of color saturation, color distribution, lightness and darkness, and keyword. If you need a picture of a house with a lot of sunset colors, you can shift over to red and orange and type “house.” Anyone who works in visual media needs to keep this one in mind, as it can be tremendously useful tool to find the picture you want not only by subject, but by appearance.

Image search engines are becoming more sophisticated, and these engines offer a mere glimpse into the future.


About the Author

Leave a comment

XHTML: You can use these html tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>