基于CSAR图像的目标高度提取方法

圆周合成孔径雷达（Circular Synthetic Aperture Radar,CSAR）可以对目标进行360°全方位观测,获得目标全方位散射信息。通过将CSAR成像数据划分成多个子孔径能够直接获取目标高度信息,本文在原有获取目标高度信息的方法基础上,进行了改进和创新,提出了一种新的提取目标高度信息的方法。本文通过对子孔径图像进行高度向投影,在一定程度上提高了计算精度,同时在子孔径选择时,也考虑了子孔径之间的相关性,有效地提高了算法的精确性。最后首次引入自主测量的L波段实测数据对本文所提算法进行验证,证实了该方法的有效性和准确性。     

基于深度学习的ViSAR多运动目标检测

视频合成孔径雷达(ViSAR)在地面动目标检测和感兴趣区域(ROI)的动态监测方面具有巨大的潜力。对地面运动目标的检测与跟踪一直是ViSAR的研究热点。针对现有基于深度学习的ViSAR动目标检测方法存在的依赖预训练模型,模型迁移难等问题,本文提出了一种基于深度学习与多目标跟踪(MOT)算法的ViSAR动目标阴影检测方法。该方法首先设计了一种从零开始深度学习的网络模型,实现动目标阴影的单帧检测。为了提高检测性能的鲁棒性,采用了基于卡尔曼滤波和逐帧数据关联的多目标跟踪算法跟踪动目标。实测数据处理结果表明该方法具有良好的检测性能。     

Bending-stretching analysis of thick functionally graded micro-plates using higher-order shear and normal deformable plate theory

In the present article, higher-order shear and normal deformable plate theory together with modified couple stress theory are developed to study the bending analysis of thick functionally graded rectangular micro-plates. One material length scale parameter is used for capturing the size effects. Utilizing the variational approach and also a principle of virtual displacement, a new form of equilibrium equations and the corresponding boundary conditions are derived. It is assumed that material properties vary through the thickness according to the power law function. Finally, an analytical solution for the bending problem of a simply supported FG rectangular micro-plate is presented.     

Nonlinear optical behavior of alkyne terminated phthalocyanines in solution and when embedded in polysulfone as thin films: Effects of aggregation

We report here for the first time, on the comparative studies of the photophysical and optical limiting behavior of our two novel phthalocyanines namely 2(3), 9(10), 16(17), 23(24)-tetrakis-(4-(5-methylhex-1-yn-3-yloxy)) phthalocyanines 2, and 2(3), 9(10), 16(17), 23(24)-tetrakis-(4-(5-methylhex-1-yn-3-yloxy)) phthalocyaninato zinc (II) (3) in solution and thin films. Nonlinear optical (NLO) properties of the samples in solution and contained in polysulfone (PSU) thin films were evaluated using Z-scan technique at 532 nm and 10 ns pulsed duration. Effects of aggregation and disaggregation on the NLO performance of metal-free phthalocyanine are reported in detail. Our findings showed that the prepared Pcs show larger nonlinear absorption coefficient and lower limiting thresholds when embedded in polysulfone films compared to in solutions. The values of nonlinear susceptibility and hyperpolarizability recorded for 3-PSU in this work are respectively, ∼2.09 × 10⁻⁸ esu and ∼1.02 × 10⁻²⁶ esu. These values are among the largest values reported for phthalocyanines and other macrocycles doped on polymer matrices.     

不同材料双层球缺罩侵彻特性的研究

采用数值模拟及钢靶侵彻试验两种方法研究了单层钛合金球缺罩、钛合金/铜、铝合金/铜双层球缺罩形成的3种杆式射流对45#钢锭靶板侵彻深度和开孔的问题。研究结果表明:不同材料的双层球缺罩形成的杆式射流对45#钢锭靶板的开孔孔径和侵彻深度大小均有直接影响,且钛合金/铜双层罩杆式射流破甲深度相对于铝合金/铜双层罩有一定提高,其开孔孔径明显增大。而单层钛合金药型罩杆式射流整体速度最大,开孔孔径较钛合金/铜、铝合金/铜双层罩杆式射流有明显提高,但破甲能力较低。研究结果对于武器战斗部设计具有一定的指导作用。     

空域反转阵导向最小方差波束形成

文章研究了基于空域反转阵（Spatial Reversion Array,SRVA）的导向最小方差（Steered Minimum Variance,STMV）波束形成方法。根据空域反转阵的原理将均匀线阵列的快拍进行空间反转,再与原阵列快拍卷积,得到孔径扩大近一倍的空域反转阵,提高阵列性能。将同样的方法用于构造空域反转阵的导向向量,实现其导向最小方差（SRVA-STMV）波束形成。仿真和海试试验结果表明,相较于常规的STMV波束形成,空域反转阵STMV波束形成能够获得更窄的主瓣、更低的旁瓣、更高的输出信干噪比以及更好的多目标分辨能力。     

基于稀疏线性调频步进信号的ISAR成像

线性调频步进信号在简化雷达系统设计的同时,也存在对高速运动目标易出现Doppler模糊的问题,因此研究如何提高其等效的重复频率具有重要意义.由于ISAR目标的后向散射场具有较强的稀疏性,即大部分能量仅由少数散射中心贡献,所以本文基于稀疏信号表示理论,通过对目标回波模型的分析,提出了一种基于稀疏步进频率信号的逆合成孔径雷达成像方法.该方法通过随机地选择线性调频步进信号的部分子脉冲进行发射,然后使用稀疏信号分解的方法对目标图像进行重建以得到目标的二维高分辨图像.该方法以计算资源为代价,能够有效地去除方位Doppler模糊,同时还能够压低旁瓣并得到超分辨的图像.仿真和实测数据处理结果验证了本文方法的有效性.     

Soil moisture estimation over vegetated terrains using multitemporal remote sensing data

A new method for retrieving soil moisture content over vegetated fields, employing multitemporal radar and optical images, is presented. It is based on the integration of the temporal series of radar data within an inversion scheme and on the correction of the vegetation effects. The retrieval algorithm uses the Bayesian maximum posterior probability and assumes the existence of a relation among the soil conditions at the different times of the series. The correction of the vegetation effects models the variation, with respect to the initial time of the series, of the component of the backscattering coefficient due to the soil characteristics as function of the variations of the measured backscattering coefficient and of the biomass. The method is tested on the data acquired throughout the SMEX02 experiment. The results show that measured and estimated soil moistures are fairly well correlated and that the performances of multitemporal retrieval algorithm are better than those obtained by employing one radar acquisition, especially in terms of capability to detect soil moisture changes. Although the approach to correct the vegetation effects on radar observations needs to be further assessed on different sets of data, this finding demonstrates that the proposed method has a potential to improve the quality of the soil moisture retrievals.     

基于Laplacian正则化最小二乘的半监督SAR目标识别

提出了一种基于核主成分分析(kernel principal component analysis,简称KPCA)和拉普拉斯正则化最小二乘(Laplacian regularized least squares,简称LapRLS)的合成孔径雷达(synthetic aperture radar,简称SAR)目标识别方法.KPCA特征提取方法不仅能够提取目标主要特征,而且有效地降低了特征维数.Laplacian正则化最小二乘分类是一种半监督学习方法,将训练集样本作为有标识样本,测试集样本作为无标识样本,在学习过程中将测试集样本包含进来以获得更高的识别率.在MSTAR实测SAR地面目标数据上进行实验,结果表明,该方法具有较高的识别率,并对目标角度间隔具有鲁棒性.与模板匹配法、支撑矢量机以及正则化最小二乘监督学习方法相比,具有更高的SAR目标识别正确率.此外,还通过实验分析了不同情况下有标识样本数目对目标识别性能的影响.     

基于多维金字塔表达和AdaBoost的高分辨率SAR图像城区场景分类算法

提出了多维金字塔表达算法, 并使用基于多维金字塔表达的AdaBoost实现了高分辨率合成孔径雷达(Synthetic aperture radar, SAR)图像的城区场景分类. 多维金字塔表达算法首先在局部特征的各维计算金字塔表达矢量, 再将所有的金字塔表达矢量连接起来构成多维金字塔表达矢量. 多维金字塔表达算法克服了金字塔表达算法在处理高维局部特征时, 遇到的输出金字塔表达矢量的区分力受计算效率制约的问题. 本文分别在一个TerraSAR-X图像库和一张大幅TerraSAR-X图像上比较基于金字塔表达的AdaBoost和基于多维金字塔表达的AdaBoost的分类性能. 实验结果表明, 与前者相比, 后者显著提高了计算效率同时保证了分类精度.