Multi-sensor Data Fusion with Dynamic Component for Context Awareness

Dempster-Shafer Evidence Theory(DST) enables multi-sensor data fusion, which makes it possible to infer the context. In this paper, we propose about how to use multi-sensor data fusion to infer context in the dynamic circumstances. Dynamic circumstance means a changing of the situation or the surrounding itself, and particularly signifies that there is a changeable factor in a specific environment. It is measures to recognize an expected situation of a moving object’s approach, contact and collision in advance. While pre-existing context inference method using Dempster-Shafer evidence theory considered belief and uncertainty at the same time, in this paper, we propose the measures that focus on change of belief. Because in dynamic circumstances, It can be an alternative of no necessity of checking uncertainty that repetitive estimating of belief in the lapse of time. With this, we can apply to recognizing a situation that rapidly proceeds as time passes.