A multi-sensor target recognition model in a complex interference environment

For multi-sensor target recognition system in a complex interference environment, the uncertainty of fused information, produced by multifold factors, leads to the fallibility of system output. This paper presents a multi-sensor target recognition model in order to solve the problem. The model revises the bias between the prior distribution and the factual distribution of target properties with the use of class confidence distance measure, and improves reliability of the target class estimate. Before fusion, the sensor information is filtrated by consistency measure and sensor confidence, which dispose conflicting information and false information effectively. The simulation results show that the model ensures accurate output of system when environment changes dynamically and the reliability of the sensors is depressed.