Nonumiformity correction algorithm of using mixed Kalman filter

A mixed Kalman filtering based nonuniformity correction algorithm is proposed,which can solve the problem of parameter drifts being nonlinear.Firstly,based on the feature of the given state-space model,the proposed algorithm employs the unscented and the traditional Kalman filtering techniques to deal with a nonlinear state transition model and a linear state observation model,respectively.Then,the estimated results can be used for nonuniformity correction.The performance of the proposed algorithm is demonstrated with simulated 1D and 2D data.Experimental results indicate that our new method extends the application range of the traditional Kalman filtering and possesses higher stability and estimation precision compared with the extended Kalman filtering algorithm.