Study Of Virtual Reality Based On Data Fusion

The paper's main work is to get explicit> simple and high efficient controlling information from the abundant data coming from multisensor in peripheral device by the theory and methods of the data fusion.The paper's innovation is it constructs a fusion-result oriented and characteristic level data fusion model. Multisensor data can be fused and inferred by this model in order to get controlling information. Multisensor system can get the information of the detected targets and its environments on high level, compared to the classical signal processing method, the multisensor data inferred by fusion method can have more complicated format and can be on different level.Fusion model can be applied to all the peripheral equipments in virtual reality system. If we use this fusion model in data processing of the multisensor data in data glove, we can do gesture recognition and interpretation. The experiment shows that the gesture recognition system based on fusion model have a higher recognition rate and better real time character.