基于广义逆高斯纹理结构的目标检测算法  

Target detection algorithm based on generalized inverse Gaussian texture structure

作　　者：陈铎 范一飞 粟嘉[1] 郭子薰 陶明亮 CHEN Duo;FAN Yifei;SU Jia;GUO Zixun;TAO Mingliang(School of Electronics and Information,Northwestern Polytechnical University,Xi’an 710072,China)

机构地区：[1]西北工业大学电子信息学院,陕西西安710072

出　　处：《系统工程与电子技术》2024年第12期4018-4025,共8页Systems Engineering and Electronics

基　　金：国家自然科学基金(62171379,62301435);中国博士后科学基金第73批面上资助项目(2023M732870);博士后创新人才支持计划(BX20230497);上海航天科技创新基金(SAST2023-044)资助课题。

摘　　要：针对海杂波模型选择失配现象以及由假定海杂波纹理结构完全非均匀导致的相干检测器性能损失问题,提出基于广义逆高斯(generalized inverse Gaussian,GIG)纹理结构空间相关性的目标检测算法。首先,分析相关杂波背景下的目标检测模型,引入海杂波纹理结构的空间相关性并将其作为先验信息;随后,基于两步广义似然比检测(generalized likelihood ratio test,GLRT)准则推导相关GIG纹理背景下的相干检测器。最后,通过理论证明所提检测器对散斑协方差矩阵和目标多普勒导向矢量具有恒虚警特性。基于仿真测试的实验结果和实测数据表明,所提检测器在均匀和部分均匀杂波背景下均提升了对海面目标的检测性能。A target detection algorithm based on generalized inverse Gaussian(GIG)texture structure spatial correlation is proposed to address the performance loss of coherent detectors caused by mismatched selection of sea clutter models and the assumption of completely non-uniform of sea clutter texture structures.Firstly,the target detection model under correlated clutter background is analyzed,and the spatial correlation of sea clutter texture structure is introduced as prior information.Then,based on the two-step generalized likelihood ratio test(GLRT)criterion,the coherent detector under correlated GIG texture background is derived.Finally,it is theoretically proved that the proposed detector has constant false alarm characteristics for speckle covariance matrix and target Doppler guidance vector.The experimental results based on simulation test and measured data show that the proposed detector improves the detection performance of sea surface targets in both uniform and partially uniform clutter backgrounds.

关 键 词：海杂波 广义逆高斯纹理 纹理空间相关性 相干检测 

分 类 号：TN911.23[电子电信—通信与信息系统]