Mobile Phone Enabled Museum Guidance with Adaptive Classific

We present a new adaptive classification system for museum guidance tasks. It uses camera-equipped mobile phones for on-device object recognition in ad-hoc sensor networks and provides location and object aware multimedia content to museum visitors. Our approach is invariant against perspective, distance and illumination. It supports the scalable identification of single objects and multiple sub-objects, pervasive tracking, phone-to-sensor and phone-to-phone communication. It adapts to user behaviour and environmental conditions over time and achieves high recognition rates under realistic conditions. Our decentralized classification approach makes the system highly scalable to an arbitrarily large number of users since the heavy-weight training process is carried out off-line on the server while the lower-weight classification task is performed individually and in parallel by each mobile phone.
Bruns, E., Brombach, B. and Bimber, O.
Mobile Phone Enabled Museum Guidance with Adaptive Classification
submitted to IEEE Computer Graphics and Applications, 2007

Category:

Science & Technology

Tags:

• computer
• augmented
• reality
• electronics
• environment
• mobile
• phoneguide