Sliding Velocity Filtering Track-Before-Detect Method With Extended Kalman Filter

Multiframe track-before-detect (MF-TBD) exhibits superior performance in weak target detection and estimation. However, traditional MF-TBD methods can only generate track fragments, rather than a continuous target track. In addition, conventional TBD methods can provide only position information, but not velocity information of a target. In this paper, a sliding velocity filtering TBD in mixed coordinates using extended Kalman filter is proposed to deal with the above problems. First, to eliminate the integration loss caused by model mismatch in range-azimuth plane, a mixed-coordinate based velocity filtering TBD is derived. This can supply polar position information and additional Cartesian velocity information, and the information increment is conducive to subsequent target tracking. The sliding window processing is applied to velocity filtering TBD to generate track fragments of different processing batches. Then, the detected points of sliding TBD, including polar position and Cartesian velocity information, are processed by extended Kalman filter to generate continuous target trajectory. Numerical simulations are conducted to demonstrate the effectiveness of the proposed method.