【摘要】：針對群目標跟蹤中的數據關聯(lián)問題,本文提出一種應用超圖匹配的多假設群目標跟蹤方法。首先將每個群作為一個整體進行跟蹤,通過引入延遲決策,利用延遲時間內產生的群航跡假設樹,對群可能發(fā)生的分離與融合行為進行判斷,實現(xiàn)對群整體的跟蹤。接著考慮群內各目標通常在運動過程中將保持相對穩(wěn)定的位置關系,應用超圖匹配算法,由航跡與量測之間的相對位置信息輔助完成近距離群內目標的數據關聯(lián)。仿真表明該多假設跟蹤方法能夠有效地對群結構進行估計。同時通過引入群內個體目標的相對位置信息,應用超圖匹配算法能夠獲得更好的群內個體目標數據關聯(lián)效果。
[Abstract]:In order to solve the problem of data association in group target tracking, a multi-hypothesis group target tracking method based on hypergraph matching is proposed in this paper. Firstly, each group is tracked as a whole. By introducing the delay decision and using the group track hypothesis tree generated in the delay time, the separation and fusion behavior of the group may occur is judged, and the whole group tracking is realized. Then, considering that each target in the group will keep a relatively stable position relationship during the moving process, the hypergraph matching algorithm is applied to complete the data association of the target in the close-range group with the relative position information between the track and the measurement. Simulation results show that the multi-hypothesis tracking method can effectively estimate the group structure. At the same time, by introducing the relative position information of the individual target in the group, the hypergraph matching algorithm can be used to obtain a better correlation effect of the individual target data in the group.
【作者單位】： 北京航空航天大學電子信息工程學院;景德鎮(zhèn)陶瓷大學機械電子工程學院;
【基金】：國家自然科學基金(61471019) 航空科學基金(20152051017) 國家留學基金(201606020013)
【分類號】：TP301.6
，

本文編號：2464814