水下目标亮点拓扑特征提取及自动识别方法

针对水下目标逆合成孔径声呐(Inverse Synthetic Aperture Sonar,ISAS)图像识别问题中观测角度随机多变,目标结构相互遮挡问题,提出一种基于多亮点拓扑矢量特征的ISAS水下目标识别方法。通过分析ISAS成像过程中散射点位置由三维空间向二维成像平面的投影关系,表明了横向定标后的声呐图像中强亮点之间的距离仅由目标散射结构之间的物理距离决定,据此基于强亮点之间的相互距离,构造能稳定描述不同观测角度下目标的拓扑矢量特征。然后通过K-means聚类获取多聚类中心以克服目标结构互相遮挡造成的亮点缺失问题。最终采用最近邻分类器实现目标识别。水池缩比模型实验表明,该方法对于水下目标的识别率达到84.0%。      

Automatic target recognition method for inverse synthetic aperture sonar imaging

To address the randomness of target aspect angle and the incompleteness of observed target in inverse synthetic aperture sonar(ISAS) imaging,a method for target recognition is proposed based on topology vector feature(TVF) of multiple highlights. Analysis of the projection relationship from 3 D space to 2 D imaging plane in ISAS indicates that the distance between two highlights in the cross-range scale calibrated image is determined by the distance between the corresponding physical scattering centers. Then, TVFs of different targets, which remain stable in various possibilities of target aspect angle, can be built. K-means clustering technique is used to effectively alleviate effect of the point missing due to incompleteness of the observed target. A nearest neighbor classifier is used to realize the target recognition. The ISAS experimental results using underwater scaled models are provided to demonstrate the effectiveness of the proposed method. A classification rate of 84.0% is reached.     

FDTD investigation on the electromagnetic scattering from a target above a randomly rough sea surface

A finite-difference time-domain approach for electromagnetic scattering characteristic from a two-dimensional (2-D) infinitely long target with arbitrary cross-section above the one-dimensional (1-D) randomly rough sea surface is presented. Taking the composite scattering of rough sea surface and infinitely long cylinder as an example, the angular distribution of the scattering cross-section with different incident angles is calculated and it shows good agreement with the numerical result by the conventional Method of Moments (MOM). Finally, the influence of the windspeed on the sea surface, the incident angle, as well as the size and location of the target on the composite radar cross-section (RCS) with different polarisations is investigated in detail.     

Deep learning-based circular disk type radar target detection in complex environment

Target identification is one of the most popular radar uses in real life. Target identification is a classifier that analyzes whether a signal contains an echo from a target (target-present) or is merely noise (target-absent). Deep learning techniques are a popular topic in classification, and they have evinced to be effective in a range of applications. In this paper, a 64 layers Circular Disk type RADAR Target Detection (CDRTD) model is proposed based on Transfer Learning using the SqueezeNet architecture of Convolutional Neural Network (CNN) that functions directly with processed radar target return eco signal and minimize the requirement of conventional laborious radar signal processing. Further, the proposed 64 layers SqueezeNet-based CNN CDRTD model was then implemented to identify circular disk type targets in complex environment. Finally, the target return eco data was tested to identify the circular disk type radar target in complex environments. We further analyzed target detection probability, false alarm rate, precision, recall, F1 in a complex environment and compared it with the ideal case. We found that our proposed CDRTD model can classify 83.3% of the test samples correctly with an overall accuracy of 94.59% in a noisy and cluttered environment whereas 100% of the test samples are classified correctly with an overall accuracy of 100% in an ideal environment.     

基于粒子群分类器的雷达目标一维像识别

设计了一种用于雷达一维像识别的粒子群分类算法.新算法首先对数据样本预处理,利用粒子群优化算法通过训练数据进行分类规则的提取,根据提取得到的规则对雷达一维像进行分类识别.基于Bayes定理和随机状态转移过程对新算法的收敛性进行分析.通过对三种飞机缩比模型的实测数据的识别实验,验证了新算法对实测数据和加噪数据均具有较高的识别率.     

基于组合支持向量机的水声目标智能识别研究

为解决水声目标小样本模式识别问题,有效地提高复杂海洋环境中的识别精度,提出了一种基于经验模式分解(EMD)、特征距离评估技术(FDET)和组合支持向量机(CSVMs)的水声目标智能识别方法。首先,将滤波、Hilbert包络解调和EMD等信号处理方法对水声目标的辐射噪声信号进行预处理,提取7个包含原始信号和预处理信号的时域和频域统计特征的特征集。然后,通过FDET从原始特征集中选择出7个敏感特征集。最后,将7个敏感特征集输入到7个支持向量机分类器中,利用遗传算法对7个分类器的结果进行合并,构成CSVMs分类器,从而实现对水声目标的智能识别。将该方法应用于舰船等水声目标的识别中,研究结果表明,该方法的识别性能优于单一SVMs分类器:同时,经过FDET得到的敏感特征集能明显地提高识别精度。     

Synthetic aperture radar image correlation by use of preprocessing for enhancement of scattering centers

We demonstrate that a significant improvement can be obtained in the recognition of complicated synthetic aperture radar images taken from the Moving and Stationary Target Acquisitions and Recognition database. These images typically have a low number of scattering centers and high noise. We first preprocess the images and the templates formed from them so that their scattering centers are enhanced. Our technique can produce high-quality performance in several correlation criteria. For realistic automatic target recognition systems, our approach should make it easy to implement optical recognition systems with binarized data for many different types of correlation filter and should have a great effect on feeding data-compressed (binarized) information into either digital or optical processors.     

A hybrid IEM–PO method for composite scattering from a PEC object above a dielectric sea surface with large windspeed: HH polarization

In the paper, for HH polarization the composite scattering from 2-D perfectly electric conducting (PEC) object situated above 1-D dielectric sea surface with large windspeed is investigated by an efficient hybrid method combining integral equation method with physical optics (IEM–PO). In the hybrid method, IEM is applied to study the direct scattering of underlying sea surface, whereas PO is employed to deal with the direct scattering of object. The mutual couplings between sea surface and object are also solved by PO method based on the equivalence principle and multi-path scattering strategy. In numerical simulations, the bistatic normalized radar cross-section (NRCS) of composite model are calculated by the proposed method and are validated by numerical experiments for different incident angles and windspeeds. The results show the hybrid IEM-PO method has a good accuracy and can greatly reduce the computational time and memory requirement.     

不同表面结构特征圆柱导体的太赫兹散射特性

太赫兹频段下导体表面的细微结构、粗糙度等细节将对目标电磁散射行为产生影响。为衡量这一影响程度,以圆柱导体为例研究了太赫兹频段下目标表面不同结构特征的电磁散射现象及其在图像域的表现规律。利用高频电磁计算方法获得了表面分别为理想光滑、带刻痕和周期粗糙的三种圆柱多姿态角、多频点单站散射场;基于转台成像算法重建了小转角下目标的二维图像。从仿真结果可以看出:m量级的细节特征在太赫兹雷达图像上有着显著的表现,表明太赫兹雷达能够获取更加丰富和精细的目标信息,从而为目标探测识别提供新的特征和技术手段。     

Performance-enhanced recognition of a far and partially occluded 3-D object by use of direct pixel-mapping in computational curving-effective integral imaging

In this paper, we propose a novel approach to enhance the recognition performance of a far and partially occluded three-dimensional (3-D) target in computational curving-effective integral imaging by using the direct pixel-mapping (DPM) method. With this scheme, the elemental image array (EIA) originally picked up from a far and partially occluded 3-D target can be converted into a new EIA just like the one virtually picked up from a target located close to the lenslet array. Due to this characteristic of DPM, resolution and quality of the reconstructed target image can be highly enhanced, which results in a significant improvement of recognition performance of a far 3-D object. In addition, the computational time required for reconstruction of a far 3-D target could be also reduced because the distance between the lenslet array and image plane is virtually shortened in the new EIA transformed by DPM. Experimental results reveal that image quality of the reconstructed target image and object recognition performance of the proposed system have been improved by 1.75 dB and 4.56% on the average in PSNR (peak-to-peak signal-to-noise ratio) and NCC (normalized correlation coefficient), respectively, compared to the conventional system.     

Bistatic occlusion effect of missile warhead based on micro-Doppler effect

Micro-Doppler effect is induced by the micro-motion dynamics of the radar target itself or any structure on the target. Spinning, coning and wobble are typical basic micro-motions of missile warhead. In this paper, a cone-shaped model for missile warhead with micro-nutation is established under the case of bistatic radar system, followed by the theoretical formula of bistatic micro-nutation is derived. Using the method of short-time Fourier transform, which confirms the simulated results are identical to the theoretical results. In addition, in order to further approach the actual environment. Base on micro-Doppler effect, the effective scattering center model with occlusion effect is considered under the case of bistatic radar system. The simulated results are compared with the fixed scattering point model without the occlusion effect. By the simulation studies, it shows the otherness between the actual environment and ideal assumption. A certain of theoretical analysis conclusions for extracting the features of missile warhead in bistatic radar system are gained, which can be applied to radar target recognition in the future work.     

基于快速轮廓转动力矩特征的激光主动成像目标识别

针对激光主动成像的图像特性,提出一种基于快速轮廓转动力矩的目标识别方法。将转动力矩的概念引入目标识别中,提出的快速轮廓转动力矩特征(FCTF)不仅包含了轮廓的尺寸、位置、规则度以及目标的亮暗等信息,同时对于旋转、尺度缩放等变换具有不变性。采用转动力矩的快速计算方法,提高了识别算法的计算效率。识别算法首先使用最大稳定极值区域(MSER)算法检测出目标特征区域,并将其变换为圆形区域,然后结合快速转动力矩特征算法提取出目标区域的局部不变特征,最后输入训练好的支持向量机分类器进行识别。实验结果表明相比于已有的激光主动成像目标识别方法,所提算法对于旋转、仿射变换均具有更高的识别率,同时单帧平均运算时间为9.68 ms,满足激光主动成像目标识别系统实时性的要求。     

不同土壤类型的粗糙地面与其下方埋藏目标复合电磁散射研究

利用MOM研究了不同土壤类型的一维有耗介质粗糙地面与其下方二维埋藏目标的复合电磁散射问题,在数值计算中,引入了锥形入射波以减少粗糙面模拟中由于人为截断所引起的计算误差.当给定入射波频率、土壤成分及土壤温度和土壤湿度后,可以根据相关模型计算出土壤的相对介电常数.数值结果以二维无限长导体圆柱为例,利用 Monte Carlo方法统计计算了双站复合散射截面随土壤类型、土壤湿度、入射波频率及均方根高度等的变化情况,并做了详细分析及讨论.该研究结果对于探地雷达在民用、军事、工业等多领域的应用具有一定的理论指导价值. 关键词： 粗糙地面 土壤类型 相对介电常数 四成分模型     

A hybrid method of solving near-zone composite eletromagnetic scattering from targets and underlying rough surface

For composite electromagnetic(EM) scattering from rough surface and target above it in near-field condition,modified shooting and bouncing ray(SBR) method and integral equation method(IEM),which are analytic methods combined with two-scale model for rough surface,are proposed to solve the composite near-field scattering problems.And the modified method is verified in effectiveness and accuracy by comparing the simulation results with measured results.Finally,the composite near-fielding scattering characteristics of a slanted plane and rough water surface below are obtained by using the proposed methods,and the dynamic tendency of composite scattering characteristics versus near-fielding distance is analyzed,which may have practical contribution to engineering programs in need of radar targets near-field characteristics under extra-low-altitude conditions.     

复杂场景下基于压缩感知的目标电磁散射与成像

研究了处于复杂场景下目标的逆合成孔径雷达(ISAR)成像问题。首先,建立了目标与复杂环境的电磁散射模型,采用计算电磁学的方法仿真得到了目标的雷达回波数据,进而充分考虑了背景噪声对雷达成像质量的影响。研究发现,目标所处的复杂背景会降低ISAR对目标的成像质量。其次,为减小仿真雷达回波数据所需的计算量,提出采用基于压缩感知(CS)的方法来对该场景进行成像,从而极大降低电磁仿真的计算点数。通过实验发现,在CS成像中,采用数据点使用率为0.4时所得到的成像质量可达到采用转台成像质量的效果。因此,采用基于CS的成像方法,可极大降低目标与场景的电磁散射计算复杂度,使得处于真实复杂场景下的目标电磁仿真和ISAR成像研究切实可行。     

Reliable approach for bistatic scattering of three-dimensional targets from underlying rough surface based on parabolic equation

A parabolic equation(PE) based method for analyzing composite scattering under an electromagnetic wave incidence at low grazing angle, which composes of three-dimensional(3-D) electrically large targets and rough surface, is presented and discussed. A superior high-order PE version is used to improve the accuracy at wider paraxial angles, and along with the alternating direction implicit(ADI) differential technique, the computational efficiency is further improved. The formula of bistatic normalized radar cross section is derived by definition and near-far field transformation. Numerical examples are given to show the validity and accuracy of the proposed approach, in which the results are compared with those of Kirchhoff approximation(KA) and moment of method(Mo M). Furthermore, the bistatic scattering properties of composite model in which the 3-D PEC targets on or above the two-dimensional Gaussian rough surfaces under the tapered wave incidence are analyzed.     

基于支持向量机的自动目标识别算法

基于对目标识别精确性的要求,提出了基于支持向量机的自动目标识别算法。介绍了基于支持向量机的自动目标识别系统的组成和识别流程,实现了目标的特征提取、SVM分类器的参数寻优,并将优化的SVM模型应用于未知图像的目标识别中。实验表明,该方法识别效果良好,具有较好的抗复杂背景的能力。     

旋翼叶片回波建模与闪烁现象机理分析

对旋翼叶片回波建模与闪烁现象进行了综合研究.基于散射点散射系数和分布情况,构建了旋翼叶片回波的散射点模型,并分析了散射点分布对回波的影响;在此基础上研究了回波时域闪烁现象的物理散射机理,并结合时频分析和横向分辨率分析了微多普勒特征及时频域闪烁现象;对两类不同分布间隔的散射点模型进行了仿真,并与积分模型进行对比性实验,结果验证了闪烁现象物理分析的合理性.该研究成果在旋翼目标的探测识别领域具有一定的理论与应用价值.     

扩展性微动目标回波模拟与特征参数提取研究

研究了微动目标的多普勒回波模拟及特征参数提取技术. 提出了一种基于物理光学法和等效电磁流法的扩展性微动目标回波模拟方法. 将在目标坐标系下计算得到的后向散射场通过坐标转换, 成为雷达坐标系下的目标回波, 通过与解析信号模型对比验证方法的正确性. 分析了圆锥与带翼弹头的进动特性, 为获得较好的时频聚集性同时避免交叉项采用S-method方法对获取的回波信号进行时频分析, 分析了不同雷达波入射角度, 不同运动状态及不同几何外形的时频分布特点. 对时频分布图进行逆Radon变换, 将正弦曲线映射到参数空间, 从而获取目标的微动参数. 该研究结合电磁散射与信号处理技术, 通过对典型弹道目标的仿真, 获得一些不同于传统微动模型的结果, 结合电磁散射理论, 对这些现象进行了解释分析. 该研究成果在弹道目标的探测识别领域具有重要的理论与应用价值.     

Novel classification rule of two-phase test sample sparse representation

In this paper, a novel classifier based on two-phase test sample sparse representation (TPTSSR) classifier and coarse k nearest neighbor (C-kNN) classifier, called novel classification rule of two-phase test sample sparse representation (NCR-TPTSSR) classifier, is proposed for image recognition. Being similar to TPTSSR classifier and C-kNN classifier, NCR-TPTSSR classifier also uses the two phases to classify the test sample. However, the classification rule of NCR-TPTSSR classifier is different to the decision rule of TPTSSR classifier and C-kNN classifier. A large number of experiments on FERET face database, AR face database, JAFFE face database and PolyU FKP database are used to evaluate the proposed algorithm. The experimental results demonstrate that the proposed method achieves better recognition rate than TPTSSR classifier, C-kNN classifier, nearest feature center (NFC) classifier, nearest feature line (NFL) classifier, nearest neighbor (NN) and so on.