Data level moving target detection algorithm based on B‐RFS. Issue 6 (1st August 2018)

• Record Type: Journal Article
• Authors: Shen, Xinglin; Song, Zhiyong; Fan, Hongqi; Fu, Qiang
• Is Part Of: IET signal processing. Volume 12:Issue 6(2018)
• Page Start: 729
• Page End: 737
• Publication Date: 2018-08-01
• Subjects: signal detection -- object detection -- random processes -- probability -- statistical testing -- state estimation; data level moving target detection algorithm -- signal‐to‐clutter ratio condition -- SCR condition -- signal level detection -- binary hypothesis -- Bernoulli random finite set detector -- B‐RFS detector -- probability -- statistical testing -- target state estimation; Signal processing -- Periodicals; 621.3822
• DOI: 10.1049/iet-spr.2017.0374
• Languages: English
• ISSNs: 1751-9675

Abstract

Abstract : Under low signal‐to‐clutter ratio (SCR) condition, the detection performance of traditional detection algorithm that detected in the signal level is poor, especially for moving the target. In this study, a new detection idea to detect in the data level is presented to improve the detection performance under low SCR condition. The new detection algorithm directly processes the measurement set generated by pre‐processing. On the basis of a binary hypothesis and the Bernoulli random finite set (B‐RFS), the authors proposed a B‐RFS detector, in which the probability of existence of target is treated as a test statistic of the binary hypothesis. As the information of multi‐frame is used, the detection performance will be better than the traditional detection algorithm. If the presence of the target is determined by the detector, they can obtain the estimation of the target state. The simulation results show that the proposed algorithm has better performance than traditional detection algorithm, and the performance of the algorithm is better with the increase of the processing time. Furthermore, the estimation of the target state is close to the true target state. As the proposed algorithm is detected in the data level, it can be applied to a variety of sensors.