(1.國防科學技術大學ATR重點實驗室,長沙 410073;2.駐廣元地區軍代室,廣元 628017)
摘要:目標編群是態勢估計的基礎。文中提出了一種面向態勢估計的多拍編群算法。首先,對目標編群問題進行了描述;其次,給出了一種緊湊群提取算法實現初始編群;最后,根據緊湊群之間的關系特征以及貝葉斯定理計算它們的合群概率,并按照定義的合群判決準則進行群的合并,得到最終的編群結果。仿真實驗證實了該算法的有效性。
關鍵詞:態勢估計; 目標編群; 貝葉斯定理; 合群判決
An Algorithm For Multi-frame Target Clustering
In Situation Assessment
Zhao Bang-xu1,2, Yang Hong-wen1
(1. ATR Key Lab., National University of Defense Technology, Changsha 410073, China; 2.Military Representation Department In Guangyuan Area, Guangyuan 628017, China)
Abstract: Target clustering is an important component in situation assessment. This paper presents an algorithm for multi-frame target clustering in situation assessment. Firstly, the problem of target clustering is described. Secondly, an extracting algorithm of compact sets is used to implement initial clustering. Finally, the clustering probability between compact sets is calculated based on their relation features and bayes theorem. And the final clustering result is obtained through combining the initial group based on the user-defined rule. The simulation has proved the effectiveness of the algorithm.
Key words: situation assessment; target clustering; bayes theorem ; combining determinant
參考文獻
[1] 韓崇昭,朱紅艷,段戰勝等. 多源信息融合[M]. 北京: 清華大學出版社, 2006.
[2] 黃雷, 郭雷. 一種面向態勢估計中分群問題的聚類方法[J]. 計算機應用, 2006, 26(5): 1109-1110.
[3] 陶午沙, 沈振康,李吉成. 基于ARG模型特征的目標陣列群識別方法[J]. 紅外與激光工程, 2004, 33(4): 371-374.
[4] 孫即祥等. 現代模式識別[M]. 長沙: 國防科技大學出版社, 2002.
[5]
張昌芳, 楊宏文, 胡衛東,郁文賢. 基于空間分布特征的陣群目標數據關聯算法[J]. 系統仿真學報, 2008, 20(22): 6074-6077.
[6] 孫亮,陳嫣,何佳洲.一種態勢估計中的目標編群方法[J].艦船電子工程,2010,30(3): 115-118.
[7] Steven M. Sassman, Northrop Grumman. Track grouping to enhance moving target exploitation [C]. Proceedings of SPIE Conference on Signal and Data Processing of Small Targets, 2000, 4048: 156-167.
[8]
李全龍, 劉洪娟, 余硌. 基于聚類思想的戰場群目標識別方法[J]. 計算機工程, 2007, 33(15): 193-195.
[9] Oliver E. Drummond, Tracking and classification with attribute data from multiple legacy sensors [C]. Proceedings of SPIE Conference on Signal and Data Processing of Small Targets, 2005, 5913: 591319-1-591319-22.
作者簡介:趙幫緒(1985-),男,黑龍江雙城人,碩士生,主要研究方向為陣群目標跟蹤、數據融合。