雷达目标识别的滤波器法

本文讨论了基于高分辨率距离像提取不敏感于方位角变化的特征矢量的方法和利用特征矢量间的相关性进行目标识别的问题，并对一些目标进行了识别实验，识别结果验证了本文所提出的利用相关滤波器法进行目标识别的有效性和可行性。     

S变换在水下目标分类识别中的应用

目标特征提取是目标识别的重要部分。介绍了一种较新的时频分析方法——S变换,对莱蒙湖底四类沉积物的反射回波进行S变换,并提出了提取变换后以频谱图的时间能量谱和奇异值为特征的特征提取方法,分析了四类回波的时间能量谱和奇异值特征的差异,并进一步用距离可分性测度检验了所提取的特征性能。最后利用最近邻分类器分类,仿真结果显示,该特征提取方法是一种有效的、稳定的特征提取方法,将在水下目标识别领域有更多的应用。     

嵌入注意力机制的卷积神经网络水声目标识别

针对低信噪比水声目标单一特征识别率低,稳健性差的问题,提出一种基于注意力机制和多尺度残差卷积神经网络（Multi-scale Residual CNN with Attention,MR-CNN-A）进行特征融合的识别方法。该方法根据多尺度卷积核与特征图形成多分辨率分析关系,并以此通过注意力机制实现优势特征权值提取与融合,从而提高模型在文中水声数据集上提取目标噪声特征和分类识别的稳健性与抗噪能力。开展了4类舰船噪声和海洋环境噪声的识别试验、水下和水面自主式水下航行器的识别试验,以及不同信噪比条件下目标噪声的识别试验。结果表明：对于文中所涉及的水声目标噪声和人工高斯白噪声干扰,该网络模型识别正确率明显高于支持矢量机与简单卷积神经网络,且对高斯白噪声的抑制能力远强于支持矢量机与简单卷积神经网络,稳健性好,模型复杂度小。     

水下目标空间源数特征提取与识别方法

针对水下目标识别正横阵位问题,并进一步扩大目标识别的阵位范围。给出了目标识别的阵列形式。用水平线列阵进行目标亮点水平方位的源数估计。提取并找出了能反映水下潜艇目标空间特征的特征量。提出了基于信噪比监督的空间源数识别方法,给出了相关矩阵的空间平滑的特征值。还提出了基于信噪比和盖尔圆半径监督的目标特征的自动分类方法。对本文提出的方法进行了大量计算机仿真,仿真结果表明。该方法可在正横阵位有效识别目标。     

基于提升小波变换的MFCC在目标识别中的应用

水下环境噪声对被动声呐目标的分类识别影响显著,为了提高水声信号识别系统对环境噪声的鲁棒性,采用提升小波变换方法提取MFCC特征,对水下被动声呐目标进行分类和识别。仿真实验表明,与传统MFCC和小波变换的MFCC相比,在信噪比相近的情况下,提升小波变换方法提取MFCC具有识别率较高、对噪声鲁棒性较好的显著优点。     

基于PCA和BP神经网络的水下目标识别方法研究

针对被动声纳信号的特点，提出了基于信号线谱特征的主成分分析(PCA)特征选择方法，其优点是从复杂的目标信号中提取目标的特有信息，降低了目标特征维数．将此方法用于实录的三类水下目标数据，采用BP神经网络对目标进行识别分类，仿真结果说明了所提出的方法的正确性和有效性．     

声呐图像背景区域灰度统计特性分析与拟合

利用声呐进行水下目标定位识别是当前水下目标识别与跟踪的重要手段之一,由于声呐图像受噪声影响严重、分辨率低,对声呐图像的背景建模有助于其目标分割与识别。首先,分析声呐图像背景区域灰度的统计特性,结合其特点采用高斯分布、Gamma分布、威布尔分布、瑞利分布模型对6类不同背景区域声呐图像统计特性进行拟合,构建声呐图像背景区域模型。最后,采用?2准则和Kolmogorov距离误差评价准则评估拟合效果。拟合结果表明,高斯分布、Gamma分布和威布尔分布均能较好地逼近声呐图像背景区灰度统计特性。为满足实时性的应用需求,选用高斯分布构建声呐图像背景灰度统计模型是可行、合理的方案,从而为声呐图像预处理和目标分割提供了背景模型建模的理论依据。     

水下激光成像系统探测距离的计算与仿真

为了估计水下激光成像系统的工作距离,根据水下激光成像系统的成像过程,通过分析目标的辐射特性,水体的衰减特性等各因素,建立了水下激光成像系统的信噪比模型.根据识别目标所需要的信噪比阈值、脉冲激光器等器件的性能指标,推导出水下激光成像系统的工作距离公式,并且完成了系统成像距离的计算与仿真.采用532 nm的Nd∶YAG固体激光器、自组ICCD相机以及基于FPGA技术设计的同步控制电路板,进行了距离选通水下激光成像实验.实验结果表明:理论模型计算的信噪比与实际图像的信噪比平均误差为1.37 dB,证实了该模型的合理性.     

Target spectra guided spectral unmixing for hyperspectral images

针对现在高光谱图像混合像元分解方法需要对所提取的端元的物理含义进行诠释的问题,提出了一种目标光谱指导下的混合像元分解方法,并给出了其具体算法实现.该方法首先针对若干给定的、具有明确物理含义的目标光谱,将光谱识别步骤引入混合像元分解过程,建立端元光谱与目标光谱间的对应关系,其次在最小距离限制的非负矩阵分解(MDC-NMF)方法基础上,引入光谱特征距离(SFD)作为正则项,以度量和保持存在对应关系的端元光谱与目标光谱间的相似性,并给出求解相应优化问题的优化算法.分别用模拟数据和真实数据对该方法的可行性和实际混合像元分解效果进行了验证,并将其与非监督情况下混合像元分解结果进行了比较分析.实验结果表明,该方法能够在目标光谱指导下较好诠释端元的物理含义,同时解决端元提取中的病态性问题.     

基于粒子群最小二乘支持向量机的前视声呐目标识别

随着声成像技术的日益发展和广泛应用,利用图像声呐进行水下目标识别逐渐成为水声探测领域的重要研究方向之一。根据前视声呐图像的特性,提出了一种水下目标识别的方法。对声呐图像进行去噪和增强处理并分割图像,来获取目标所在区域、提取目标的区域形状特征;利用粒子群算法优化最小二乘支持向量机的正则化参数和核参数,构造出高性能的多分类器;输入待识别目标的特征实现分类。实验表明:优化后的最小二乘支持向量机能够准确、有效地识别出水下目标,并且具有较高的精度。     

基于随机森林的统计控制图模式识别研究

引入随机森林方法进行统计控制图模式识别的研究。提取了控制图的统计特征和形状特征,设计了5种不同的特征组合方法,利用蒙特卡洛仿真方法产生训练数据集和测试数据集,选取了常用的3种模式识别方法（支持向量机方法、人工神经网络方法、决策树方法）进行对比。实验结果表明,随机森林方法相比其他3种分类器方法,在分类准确率和消耗时间两个维度上都有明显优势,可以应用于统计过程控制图模式识别。     

基于边界不变特征的扩展目标识别方法

提出一种新的以边界不变矩作为识别特征,运用BP网络识别扩展目标的方法。首次通过详细的理论证明和实验分析,揭示了离散边界不变矩不再具有严格的比例不变性,而位移和旋转不变性保持相对稳定,并对该不变矩作为识别特征的误差进行了深入分析,给出了正确计算边界不变矩的途径。在此基础上,以该边界不变矩作为识别特征,输入BP网络,采用合理的网络结构,实现对发生位移、旋转和尺度变化的扩展目标的识别。边界不变特征的引入,减少了数据运算量,实验结果表明,识别率达到95.9%。     

Face recognition using Krawtchouk moment

Feature extraction is one of the important tasks in face recognition. Moments are widely used feature extractor due to their superior discriminatory power and geometrical invariance. Moments generally capture the global features of the image. This paper proposes Krawtchouk moment for feature extraction in face recognition system, which has the ability to extract local features from any region of interest. Krawtchouk moment is used to extract both local features and global features of the face. The extracted features are fused using summed normalized distance strategy. Nearest neighbour classifier is employed to classify the faces. The proposed method is tested using ORL and Yale databases. Experimental results show that the proposed method is able to recognize images correctly, even if the images are corrupted with noise and possess change in facial expression and tilt.     

一种模糊红外目标边缘和不变特征提取方法

针对红外图像中模糊目标难以分割和识别的情况,提出以小波多尺度滤波分解后的逼近图像作为自适应域值进行图像二值化的方法,由此分割出模糊图像中的目标,提取目标的轮廓边缘.然后,提出一种新的矩计算方法提取目标的不变性特征.实验结果表明,该方法具有很强的抗噪和抗扰性能,能有效提取复杂背景中模糊目标的平移、缩放和旋转不变量,极大提高了运动模糊目标识别的可靠性.     

Detection for high resolution radars

Abstract

In this paper we discuss detection problems for a high resolution radar. Fluctuation in the target radar cross section usually decreases the probability of detection. However, through integration of cells within range profiles of a high resolution radar, variation of the integrated magnitude with respect to the change of carrier frequency and target aspect becomes much smaller, and this is helpful for improving the probability of detection. Two detection algorithms, the cell integration method and the correlation method, for a high resolution radar are proposed, and their detection performances are compared with that obtained by a conventional low resolution radar. Some theoretical formulations are developed. Simulation results show the effectiveness of the proposed algorithms.     

Ground-based differential absorption lidar for water-vapor profiling: assessment of accuracy, resolution, and meteorological applications

The accuracy and the resolution of water-vapor measurements by use of the ground-based differential absorption lidar (DIAL) system of the Max-Planck-Institute (MPI) are determined. A theoretical analysis, intercomparisons with radiosondes, and measurements in high-altitude clouds allow the conclusion that, with the MPI DIAL system, water-vapor measurements with a systematic error of <5% in the whole troposphere can be performed. Special emphasis is laid on the outstanding daytime and nighttime performance of the DIAL system in the lower troposphere. With a time resolution of 1 min the statistical error varies between 0.05 g/m(3) in the near range using 75 m and-depending on the meteorological conditions-approximately 0.25 g/m(3) at 2 km using 150-m vertical resolution. When the eddy correlation method is applied, this accuracy and resolution are sufficient to determine water-vapor flux profiles in the convective boundary layer with a statistical error of <10% in each data point to approximately 1700 m. The results have contributed to the fact that the DIAL method has finally won recognition as an excellent tool for tropospheric research, in particular for boundary layer research and as a calibration standard for radiosondes and satellites.     

Motion feature extraction for stepped frequency radar based on Hough transform

The high resolution range profile can be obtained through coherent pulse compression processing in every coherent processing interval (CPI) for stepped frequency radars (SFR). The radar observation time can be divided into multiple CPIs. By utilising the temporal information inherent to a sequence of these range profiles, a time-range image containing the target motion information can be synthesised. Through analyses of the point scatterer targets with the radial uniform motion, radial uniformly accelerated motion, rotation and vibration, this study indicates that there is a certain correlation between the target motion features and the time-range image features in the limited observation time. Based on this result, we propose a motion feature extraction method for SFR without any assistant information using the conventional and well-known feature extraction tool, Hough transform. The authors give detailed analyses and explanation of this method, and verify them by outfield experiments, demonstrate the motion feature extraction results using real radar data. The proposed method enables the motion feature extraction through space non-coherent integration, which fully utilises the radar data information under low signal-to-noise ratio conditions.     

Statistical analysis of the zero-phase method for aligning noisy high-resolution radar signals

Automatic target recognition using high-range resolution radars is a difficult task. Variations in the distance to the target cause circular shifts of the received signal, and most of the classification algorithms are very sensitive to shifts over the input signal. This fact makes the alignment of each signal prior to any classification stage very important. Here the alignment of noisy signals using the zero phase method is studied. In order to evaluate the performance of the alignment method, a theoretical analysis of the sensitivity to noise of this alignment method is carried out. As a result, an analytical expression that predicts the error of the alignment method is obtained. The validity of this expression is also confirmed by experimental results. A database of high-range resolution radar profiles containing patterns belonging to six different targets has been used, and a comparative study of the sensitivity to noise estimated using the profiles of the database and predicted by the analytical expression is carried out. The results demonstrate that the proposed analytical expression is useful to analyse the sensitivity to noise of the zero-phase alignment method for medium and high signal-to-noise ratio values.