Ocutag, a division of Ricoh Innovations, aims to help apps deliver content to users taking photos. Their aim is to herald a new model of retail. App designers upload images and tag them with information. When users snap pics of the same thing in reality, information can be delivered through the app, such as that appearing in the graphic of the scarf below: where to buy it, details about it, local stores that carry it. What information is delivered is up to the company with the app. The company provides the Ocutag API for integration.
The Ocutag API is REST and works the system like a loop: the user takes a photo of something that the app can recognize. Examples given are Bollywood stars in India and items in a shop. Of particular use might be more information about a food item when you are grocery shopping: where was it grown, for example? The image is automatically uploaded to the Ocutag server where any objects in the photo are recognized as being part of the app’s collection. Then information on those apps is sent back down to the user. In the graphic, only one object, the scarf, is recognized. But Ocutag has examples where multiple objects are recognized, with information delivered on each one.
As the Ocutag API documentation specifies, its image recognition service uses just 3 request-response pairs:
“The Ocutag web service uses two sets of APIs: the Collection API and the Store API. Within the Collections API, there are two calls: one to get a list of collections, and one to access the image recognition service. Within the Store API, there’s only one call: to retrieve any metadata associated with the recognized image.”
The nifty thing about this is convenience: users don’t have to worry about scanning barcodes. Pricing has not been set as of this writing but a free trial is available.