Google Labs is testing a new image search feature by using something called the Image ‘Swirl’. Google Image Swirl integrates the Similar Image Search and Google’s Wagon Wheel. The Image Swirl is set to simplify your search without needing any customisation to the search query, like changing the search term to relate better to what you are trying to search or to customise the query to get the desired image resolution.
The image swirl shows a cluster of search results in a 4 x 3 matrix form. Each of these 12search results groups similar images into one. The search results look like a pile of images assorted in twelve parts (obviously according to the relevance). When you click on any of these image piles, a “Wonder Wheel” is opened that contains a more tightly refined search result.
Google Image Swirl hands on:
We tried the Google Image search for ‘Apple’. The search results showed the 4×3 matrix of results, which contained the Apple fruit, the Apple iPhone, Apple Store and a few other collections of related searches.
After clicking on the iPhone image, a more customised search results appeared in the a Wagon Wheel showing similar images of the iPhone. Interestingly, these image results are scaled down versions of their actual resolution. So if we want a high resolution image, we simply click on the bigger image result, which then takes us to the webpage where the image is located.
Google Image Swirl is an interesting tool but still lags behind the decision support feature of Bing’s Visual Search. Image Swirl currently supports only 200,000 search terms with the auto-complete support. Google is planning to add more search terms to the list with the release of the Image Swirl.
Overall, Image Swirl is a fancy way of searching Google Images, with reduced customisation efforts. We would be more interested if Google integrates a decision support system like Bing’s Visual Search into their Image Swirl.