Efficient fusion technique for disparate sensory data

The author proposes a general methodology for multisensor data fusion. The sensory data may be partial or indirect. This method introduces the definition of three primitive sensory data types and provides general, sensor-independent, and practical solutions for fusion of different types of data. For a multisensor system, this method can deal with disparate sensory measurements for position or relationship, implement integration of derived information in an efficient manner, and give more accurate estimates. A case study and Monte Carlo simulations illustrate the application of the proposed method to greatly improve the position and orientation estimation for a mobile robot and show the statistical effect of two-dimensional data integration