On Improved Bernoulli Filter for Extended Target Tracking

The extended target tracking (ETT) is widely used in various surveillance applications. In order to improve tracking accuracy and filtering efficiency of the existing Bernoulli filters, we present an improved Bernoulli filter in this paper. Firstly, the filtering equations of the proposed filter are analytically derived. After employing the weight optimization scheme, we discuss the particle implementation based on adaptive detection gate and scatter number. Finally, the numerical simulations indicate that the proposed filter can achieve the ETT with promising tracking performance.