26.01.12

Recognition system identifies users by their shoes

A research team in Germany is using a Kinect based system to recognise users of interactive tabletop displays by their shoes. The Bootstrapper project, led by Prof. Patrick Baudisch at the Hasso Plattner Institut (HPI) in Potsdam, Germany, can provide users of tabletop computers with personalised functionality and log their activities and progress. The institute has created a video showing Bootstrapper put to use in an education environment.

Bootstrapper uses one or more depth cameras to observe the shoes of users interacting with the interactive table display, such as Microsoft Surface.

It then identifies users by matching camera images with a database of known shoe images. When multiple users interact, Bootstrapper associates touches with shoes based on hand orientation.

Baudisch claims the approach can be implemented using consumer depth cameras because shoes offer large distinct features such as colour and they shoes naturally align themselves with the ground, giving the system a well-defined perspective.

A study into the effectiveness of Bootstrapper reported that the system recognised participants from a database of 18 users with 89% accuracy, based on a single observed frame.

However, the power of Bootstrapper is diminished when users change their shoes or multiple users wear the same shoes.