Efficient and Enhanced Multi-Target Tracking with Doppler Measurements

摘要: In many radar and sonar tracking systems, the target state typically includes position velocity components that are estimated from a time sequence of Doppler measurements. The use measured information directly in trajectory estimation leads to nonlinear filter implementation such as extended Kalman (EKF), particle etc. We investigate method for including measurements part data association process then assess benefits this approach. It is well understood performance can dominate total tracker designed track targets presence clutter. case, component may be used combination with measurement an additional discriminant origin. have developed simple but efficient (DDA) which utilises both single multi-target tracking. If not estimation, filters incorporation required, however significant improvement still observed. proposed DDA demonstrated using linear integrated probabilistic algorithm (LMIPDA) splitting (LMITS) active underwater scenario.