Google has debuted their new visual search for Google Images (images.google.com). It looks a lot like the image search function in Microsoft’s Bing. Rather than a rip-off, though, Google’s search goes a step further than Bing and includes a lot of useful information that goes along with each image.
From the index of over 10 billion images in Google’s databases, Google has cross-referenced images as often as possible in order to more thoroughly identify what’s in the picture. For instance, if you post a photo of The Sands Hotel in Las Vegas, but do not label it (even with a descriptive file name), Google will likely figure out what it is by comparing it to other photos of the hotel that have been labeled. This is also true of many other types of images, even down to sub-species of animals.
When photos are displayed on the Google search, they pop up as a box over the top of the original page they’re on and this box has other information about the picture. It tells you the size, weight (file size), any image information embedded (titles, photographer, usage rights, etc.) and gives you its best-guess if there is no labeling. You can also use that image as the basis for a new search to find more like it.
You can even specify a color you prefer when searching for an image in the Advanced Search area. So if you’re looking for Volkswagen Jettas, but only want to look at red ones, you can include that and Google will do its best to only return red VW Jettas.
All in all, there are a lot of cool new things in Google’s new Image Search. It’s now become on-par with Bing and then one-upped the competition.