Digital Diorama: Sensing-based real-world visualization


Many sensors around the world are consistently collecting the real-time real-world data. The data streams captured by these sensors can give us an idea of what is going on in a specific area; however, it is not easy for humans to understand their spatial and temporal relationships by just looking at them independently. This paper proposes to construct Digital Diorama, a three-dimensional view where viewers can see at a glance how people are moving around the monitored space without violating their privacy, by integrating multiple data streams captured by stationary cameras and RFID readers in real time. Digital Diorama realizes such real-world visualization with the following features: 1) view control, 2) real-time camera image superimposition, and 3) privacy control. We have demonstrated that Digital Diorama for a shopping center was able to present the current positions of persons and real-time camera images in approximately 1 frame per second.

Proc. International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems (IPMU)