Mobile Object Recognition for Retail

Mobile Object Recognition for Retail

Take a snapshot with your iPhone of a supermarket’s flyer and receive information and discounts on the products photographed, or put them into a shopping cart. It can’t be simpler than that for an end-user, can it? The steps involved in setting up this project are:

  1. Collecting digital images of the products to be recognized
  2. Tagging these images with metadata and putting them into a special feature-point database
  3. Loading the database onto a server
  4. Implementing an object recognition algorithm on the server
  5. Building a mobile front-end that communicates with the server

After having implemented numerous applications using object recognition on various different channels (mobile devices, computer applications, web services, web pages) we realized a number of things :

  • Each new project only required it’s own front-end (step 5) and one day of work for steps 1-4
  • Steps 1 and 2 require software built in-house, there are no products on the market that do this
  • Steps 3 and 4 never had to get redone, simple copying the code from a previous project was enough
  • Step 5, the front-end, can be done by a team of developers with no prior knowledge of object recognition

So it happened that this project was put together in just a few days since the image collection and database building steps had already been build with the Automated Visual Repository Building project and setting up the object recognition code and the servers had been done during the Pocket Supercomputer project.

Scientific Advisor

Our role for this project, besides providing the database back-end, was as a scientific and technical advisor. At first instance the SIFT object recognition algorithm was used to recognize the products, but at a later stage the SURF algoritm–part of Intel’s opencv library–was tested to replace the good, but too expensive SIFT algorithm.

Leave a Comment