航空学报 > 2024, Vol. 45 Issue (12): 329466-329466   doi: 10.7527/S1000-6893.2023.29466

基于TransUNet的机动微弱目标智能检测前跟踪算法

武星蕊1, 彭锐晖1,2(), 孙殿星1, 谭顺成3, 张一泓1, 韦文斌1   

  1. 1.哈尔滨工程大学 青岛创新发展中心,青岛 266000
    2.哈尔滨工程大学 信息与通信工程学院,哈尔滨 150001
    3.海军航空大学 信息融合研究所,烟台 264001
  • 收稿日期:2023-08-22 修回日期:2023-09-22 接受日期:2023-10-28 出版日期:2023-11-23 发布日期:2023-11-22
  • 通讯作者: 彭锐晖 E-mail:pengruihui@hrbeu.edu.cn
  • 基金资助:
    山东省自然科学基金面上项目(ZR2020MF015);航天科技集团稳定支持项目(ZY0110020009)

Intelligent TBD algorithm for maneuvering weak targets based on TransUNet

Xingrui WU1, Ruihui PENG1,2(), Dianxing SUN1, Shuncheng TAN3, Yihong ZHANG1, Wenbin WEI1   

  1. 1.Qingdao Innovation and Development Center,Harbin Engineering University,Qingdao 266000,China
    2.College of Information and Communication Engineering,Harbin Engineering University,Harbin 150001,China
    3.Information Fusion Research Institute,Naval Aviation University,Yantai 264001,China
  • Received:2023-08-22 Revised:2023-09-22 Accepted:2023-10-28 Online:2023-11-23 Published:2023-11-22
  • Contact: Ruihui PENG E-mail:pengruihui@hrbeu.edu.cn
  • Supported by:
    Natural Science Foundation of Shandong Province(ZR2020MF015);Aerospace Technology Group Stability Support Project(ZY0110020009)

摘要:

隐身技术的发展和运用给雷达目标探测带来了极大的挑战,检测前跟踪技术作为微弱目标探测的重要技术方向之一,常被用于雷达探测中,但传统的检测前跟踪算法计算量大、受目标运动模型限制,难以有效检测各种不同运动状态的机动目标。针对该问题,提出基于TransUNet的微弱目标智能检测前跟踪算法,算法不受任何运动模型限制,可以有效识别各种运动状态的机动目标。首先,利用目标运动轨迹的空-时相关性特征,基于TransUNet模型设计了一种适用于不同运动状态的机动微弱目标粗检测算法,实现微弱机动目标的初步检测;然后,利用初步检测结果对目标量测进行区域搜索,实现微弱机动目标点迹的回溯;最后,针对低信噪比情况下初步回溯点迹的漏检问题,利用部分成功检测点迹的辅助信息,结合最佳抛物线值函数实现目标点迹的精细检测和航迹的完整回溯。仿真试验验证了本方法能够对多种运动状态的微弱机动目标实现有效检测。

关键词: 检测前跟踪, 微弱目标, TransUNet网络, 空-时相关性特征, 目标检测, 最佳抛物线值函数

Abstract:

The development and application of stealth technology have brought great challenges to radar target detection. As one of the important technical directions for weak target detection, the Tracking-Before-Detection (TBD) technology is often used in radar detection. However, the traditional TBD requires a large amount of computation and is limited by target motion models, making it difficult to effectively detect maneuvering targets in various motion states. To address this issue, a TransUNet based intelligent TBD algorithm for weak targets is proposed. The algorithm is not limited by any motion model and can effectively identify maneuvering targets in various motion states. Firstly, utilizing the spatio-temporal correlation characteristics of the target motion trajectory, a rough detection algorithm for maneuvering weak targets in different motion states is designed based on the TransUNet model to achieve preliminary detection of weak maneuvering targets. Then, the preliminary detection results are used to perform regional search for the target measurement, and achieve backtracking of weak maneuvering target points. Finally, to address the problem of missed detection in preliminary backtracking points at low Signal-to-Noise Ratio (SNR), the auxiliary information of partially successfully detected points, combined with the Optimal Parabolic (OP) value function, is utilized to achieve precise detection of target points and complete backtracking of tracks. Simulation experiments have verified that this method proposed can effectively detect weak maneuvering targets in various motion states.

Key words: Tracking-Before-Detection (TBD), weak target, TransUNet network, spatio-temporal correlation characteristics, target detection, optimal parabolic value function

中图分类号: