基于矩阵分析的加权证据时空融合算法研究

在多传感器目标识别系统中．为解决各传感器可信度不一致的问题，提出加权证据的方法。为便于采用计算机实现，采用基于矩阵分析的D-S证据理论。考虑融合系统的时间性和空间性．先从空间域对多传感器进行空域信息融合．然后在此基础上对时间序列进行融合。最后进行判决，识别出目标。识别实例表明了该方法的正确性和有效性。

Research of Temporal-spatial Fusion Algorithm Based on Weighted Information and Matrix Analysis

In order to solve the different reliability of every sensor, the method of evidence weight is introduced in the multi-sensors system of target identification. Considering the temporal character and spatial character of the fusion system sufficiently, the D-S evidence theory based on matrix analysis is introduced. This is helpful for computing. The information is firstly fused in space, and then the result is fused in time order. The target is recognized by final estimation. An identification example using this method indicates that the precision and effectiveness are raised.