Acta Optica Sinica, Volume. 31, Issue 2, 228002(2011)
An Algorithm of Cluster Tracking for Midcourse Ballistic Object Group by Infrared Multi-Sensor Based on Probability Hypothesis Density Filtering
The essentiality of cluster tracking for midcourse ballistic target group by low earth orbit(LEO) optical constellation is analyzed. And an improved method of cluster tracking is presented. The probability hypothesis density filter is used to track multi-target on the focal plane, filter out clutter, and estimate target number and state. Then for improving the stability of focal plane target tracking, an image registration method is adopted to adjust interframes target state estimation. Finally, the focal plane 2D centroid measurement of target group is computed based on the focal plane tracking results, and unscented Kalman filter(UKF) is adopted to track the 3D centroid of target group through multi-sensor fusion in sequence. Simulation results show that, the proposed method works well under different clutter rates, and gains a better precision than the traditional one. Meanwhile, it successfully realizes the tracking of each target on the space-based infrared focal plane.
Get Citation
Copy Citation Text
Lin Liangkui, Xu Hui, Long Yunli, An Wei, Xie Kai. An Algorithm of Cluster Tracking for Midcourse Ballistic Object Group by Infrared Multi-Sensor Based on Probability Hypothesis Density Filtering[J]. Acta Optica Sinica, 2011, 31(2): 228002
Category: Remote Sensing and Sensors
Received: Jun. 23, 2010
Accepted: --
Published Online: Jan. 30, 2011
The Author Email: Liangkui Lin (kk2buaa@163.com)