高速动目标多频段融合一维超分辨成像算法研究  

One-dimension super-resolution imaging algorithm with high-speed moving-target multi-frequency fusion

作　　者：李威[1] 梁福来[1] 张毅[2] 吕昊[1] 罗二平[1] 

机构地区：[1]第四军医大学生物医学工程学院,西安710032 [2]第四军医大学训练部教保处,西安710032

出　　处：《医疗卫生装备》2015年第10期1-7,共7页Chinese Medical Equipment Journal

基　　金：陕西省社发攻关项目(2014K11-04-01)

摘　　要：目的:提出一种新的融合成像算法,解决现有目标模型及算法在宽带条件下对高速动目标频响建模不精确的问题,提高成像分辨率。方法:将衍射几何理论(geometrical theory of diffraction,GTD)模型与目标运动相结合,推导出宽带雷达动目标回波频谱模型,分析动目标多频段雷达回波间的非相干量,在频域对每部雷达的回波谱进行匹配滤波和速度补偿,基于实包络对齐估计线性相位项,利用旋转矢量不变技术(estimating signal parameters viarotational invariance techniques,ESPRIT)算法特性直接估计散射中心的绕射系数和位置,最后估计散射强度。结果:通过仿真实验,新的图像融合算法与传统算法相比,在线性相位项的补偿和强散射中心个数估计的精度上均有所提升,成像效果显著改善,且新算法大大地精简了计算过程。结论:新的融合成像算法可显著提高快速运动目标距离像的分辨率。Objective To propose a new image fusion algorithm to improve the imaging resolution by enhancing the precsion of the existing target model and algorithm for modeling high-speed moving-target frequency response in broadand condition. Methods Geometrical theory of diffraction（GTD） model was combined with target motion to derive a wideband radar dynamic model of target echo spectrum, and the non coherent volume of the moving-target multi-frequency radar echos were analyzed. Matched filtering and velocity compensation were carried out for the echo spectrum of each radar in the frequency domain. The linear phase term was estimated based on real envolop alignment, and the diffraction coefficient and site of scattering center were determined by using estimating signal parameters viarotational invariance techniques（ESPRIT）, then the scattering intensity was aslo appraised. Results Simulation experiments showed that the new image fusion algorithm behaved better than the traditional one in compensation of linear phase term, estimation precision of strong scattering centers number, imaging effect and computational procedure. Conclusion The new image fusion algorithm may enhance the resolution of the fast moving target range image.

关 键 词：高分辨距离像 线性频率调制 速度补偿 相干融合 

分 类 号：TN957.52[电子电信—信号与信息处理]