A Novel Adaptive Extended Target Tracking Algorithm Based-on Variational Bayes Within Unknown Probability of Measurement Loss

A novel variational Bayes based adaptive extended target tracking algorithm is proposed to solve the problem of extended target tracking with an unknown probability of measurement loss. Firstly, in the random matrix theoretical framework, an unknown positive semi-definite matrix represents the range of the target. In addition, the rotation matrix represents the change of the target direction. Secondly, the Bernoulli random variable with binary sequence properties is used to determine whether measurement loss occurs. Thirdly, The Beta distribution is chosen as the prior distribution for the probability of measurement loss. Lastly, the kinematic state, the extended information and the probability of measurement loss are jointly estimated based on the VB approach.