Snooth just launched an iPhone application which uses image recognition to do something genuinely useful (as least to me). If you take a picture of a wine label, Snooth finds the stores closest to you that stock that wine and the prices in each store. You can also read reviews of the wine and find similar bottles by winery, region or varietal.
The app uses a specialized image-recognition technology from Canadian startup TinEye. Most wine labels are on a curved surface, so the software must be able to find images that are slightly rotated, warped, cropped, blurred, have skewed colors or have been taken in a location with low lighting like a restaurant. In fact, identifying where to buy that wonderful house wine is one of the main scenarios in which you might use Snooth’s app. Others include price comparison when you are in a wine shop.
820,000 wines are searchable via image recognition, and Snooth links to 11,000 wine merchants worldwide. Snooth is not the first company to consider this application for image recognition. The most notable is Google Goggles, but the Goggles database never actually listed a useful number of labels.
Snooth Wine Pro is available in the iTunes App Store for $4.99. Users can also install Snooth Wine, a free, ad-supported version. Snooth’s core product is a social, wine review site whose business model is based on lead generation and selling data to wineries and wine merchants on where potential customers might be located. Snooth is based in New York and is privately funded.
Getting content noticed is a challenge for everyone making apps. We’ll cover the topic at DiscoveryBeat 2010. Startups and big companies alike are encouraged to submit their discovery tactics to our Needle in the Haystack competition. Early bird discounts are available until September 22. Sponsors can contact us at firstname.lastname@example.org. To buy tickets, click here.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.