Improved multitarget track-before-detect for image measurements

Track of weak targets with unknown and time varying number is an important and difficult issue. In the paper, an improved multitarget track-before-detect (TBD) method for image measurements is proposed to tackle this problem. The method is developed in the framework of probability hypothesis density (PHD) filtering by utilizing the independent measurements from multiple homogeneous sensors. To fast implement the PHD based method using particle filtering technique, an efficient implementation approach is also presented by partitioning the particles into multiple subsets. Simulation results show the improved performance of the proposed method for multitarget TBD at low signal-to-noise ratio (SNR) condition.