Multi-sensor target tracking based on Gauss-Markove estimate

As measurements errors are assumed to be zero-mean, Gaussian distributed and uncorrelated, an improved method of target state estimation for multi-sensor target tracking is presented based on Gauss-Markove estimate. By minimizing the error covariance matrix, weighted least square method is exploited to obtain the Gauss-Markove estimate. Applying Gauss-Markove estimate to Kalman filter, the target state was estimated. The Monte-Carlo simulation results show that Gauss-Markove estimate can efficiently improve the accuracy of observations. The performance of proposed algorithm for state estimation is better than that with single sensor observations. Presented approach and obtained results may be useful in multi-sensor target tracking.