海杂波背景下小目标检测的分形方法

针对海杂波背景下小目标检测对海情依赖性强的问题, 本文采用分数布朗运动模型对实测海杂波建模, 结合多重分形去势波动分析法确定分形参数, 分析了海杂波的单尺度、多重分形特性. 在单尺度分形的基础上, 利用表征海杂波分形特征的分数维和Hurst指数构建了分形差量, 提出了基于分形差量的小目标检测方法;在多重分形基础上, 比较了两种海杂波的高尺度多重分形特性. 结果表明, 当尺度q > 10时, 纯海杂波的多重分形参数H(q) < 0, 而存在小目标的H(q) > 0, 此差异性为高尺度分形参数的海杂波背景小目标检测提供了判定依据. 所研究的两种方法均能实现不同海情下的小目标检测.     

基于多重分形谱的高光谱数据特征提取

针对单一分形维数在高光谱数据处理中的不足,提出了一种基于多重分形谱的光谱信号奇异性特征提取方法,引入多重分形谱表征光谱曲线的奇异性特征.该方法根据分形测度将光谱曲线进行划分,用光谱概率测度计算配分函数,通过尺度指数的Legendre变换实现光谱曲线多重分形谱的提取,根据各类地物间的类别可分性准则Bhattacharyya距离选择有效特征,最后利用地物分类实验来验证该方法的有效性.实验结果表明,多重分形谱用于分类时分类精度达95.2%,当其维数为原数据波段数的10%时,总体分类精度仍可达82.2 %.多重分形谱表征了具有相同奇异性的波段子集的分形维数,准确的描述了光谱曲线的奇异性和分布特点,该方法能够有效地实现高光谱数据的特征提取.     

海杂波背景下的目标检测新方法

提出了一种基于分形布朗运动模型的S波段雷达海杂波分形维数提取方法.分析了基于记忆库混沌时间序列预测方法，引入一种改进核函数的支持向量机分类器.在此基础上，提出了一种新的海杂波背景下目标检测方法.应用S波段雷达实测海杂波数据，计算得到了该信号的分形维数与Lyapunov指数，验证了S波段雷达海杂波的混沌分形特性.仿真实验结果验证了该方法具有较强的检测能力和抗杂波性能. 关键词： 分形布朗运动 分形维数 记忆库预测方法 支持向量机分类器     

基于双树复小波变换的非平稳时间序列去趋势波动分析方法

多重分形去趋势波动分析是研究非平稳时间序列非均匀性和奇异性的有效工具, 针对该方法中趋势项难以确定的问题, 提出一种基于双树复小波变换的方法, 实现了非平稳信号的多重分形自适应去趋势波动分析. 利用双树复小波变换提取信号的多尺度趋势和波动信息, 通过小波系数的希尔伯特变换确定每个时间尺度不重叠子区间的长度, 使多重分形分析具有信号自适应性及较高的计算效率. 以具有解析形式分形特征的倍增级联信号和分数布朗运动时间序列为例验证本文方法的有效性, 所得结果与解析解相吻合. 与传统的多项式去趋势多重分形方法相比, 本文方法根据信号自身特点自适应地确定信号的趋势和不重叠等长度子区间长度, 所得结果更加精确. 对倍增级联信号时间序列取不同的长度, 验证了算法的稳定性. 分别与基于极大重叠离散小波变换和离散小波变换多重分形方法进行比较, 表明本文方法具有更精确的结果和更快的运算速度.     

海杂波的分段分数布朗运动模型

主要研究了分段分数布朗运动(PFBM)模型在雷达海杂波分形建模中的应用.由于自然界和人造系统中研究对象不具有数学上完美的分形特性, 从而研究对象的分形特性无法在整个尺度区间上成立, 传统上, 海杂波的单一分形模型仅利用无标度区间内海杂波的自相似信息进行参数估计, 并没有考虑海杂波在无标度区间以外的尺度下所包含的信息.分段分数布朗运动从频域角度对海杂波频谱进行分段描述, 对应到时域即从粗略尺度和精细尺度两方面描述海杂波时间序列.结合海杂波产生的物理背景, 该模型可以为海杂波时间序列在粗略尺度和精细尺度下表现出的不同粗糙度提供机理性解释.在此基础上, 还研究了具有不同多普勒频率的运动目标对海杂波的影响, 结果表明运动目标对粗略尺度和精细尺度下海杂波的影响程度是不同的.     

海杂波FRFT域的分形特征分析及小目标检测方法

针对海杂波背景下海情对小目标检测的严重影响, 本文研究了实测海杂波在分数阶Fourier变换(FRFT)域的分形特征, 分别提出了单、高尺度下的分形检测方法. 由数学定义推得, FRFT 在不同阶数和尺度情况下, 不具有一致的自相似特性, 采用多重分形趋势波动分析法确定分形参数H(q), 分析了海杂波在不同海情、距离和极化条件下的分形特征. 在单尺度基础上结合FRFT的变阶优势, 提出了阶数自适应的小目标检测方法; 高尺度条件下, 比较了不同因素对海杂波FRFT域多重分形参数的影响. 结果表明:海杂波FRFT域可用变换阶数的方法检测到湮没在复杂海情中的小信号, 检测门限多数提高200%以上, 比采用时域信号提高26.3%. H(q) 在负高尺度上具有明显的多重分形特征差异, H(q)-q曲线满足反正切分布, 纯海杂波与含目标数据的拟合幅值比分别大于1.8(HH)和1.4(VV), 为海杂波背景小目标检测提供了判定依据.     

多重分形分析图象边缘提取算法

提出了一种基于多重分形分析的图象边缘提取算法,通过计算每个象素点的奇异值和多重分形谱,并根据多重分形谱的各种测度修正,提取出图象的边缘信息.在分析图象的各种象素点的多重分形谱特性的基础上,着重分析了多重分形谱常用的若干测度以及选取标准.该算法利用奇异性Hölder指数和多重分形谱以及它们组成的判据来进行图象边缘提取的思路不同于传统的基于梯度的局部极值点来进行图象边缘提取的方法.实验结果表明:该算法可以在保留重要边缘信息的情况下去除不重要细节,更能符合人的视觉心理.     

气液两相流波动信号多重分形去趋势波动分析

为了研究不同流型压差波动信号的分形特征,本文首先讨论了多重分形去趋势波动分析对周期信号、随机信号和混沌信号的识别能力,然后运用多重分形去趋势波动分析方法对水平管内气-液两相流压差波动信号进行分析.通过计算广义Hurst指数、尺度函数,多重分形谱,细致量化了压差波动信号的局部及不同层次的波动奇异性.研究结果表明:压差波动...     

时频域分形维数分析的光谱信号重叠峰解析算法

由于光谱谱线存在自然展宽、多普勒展宽、碰撞展宽等,使混合气体中多种成分的吸收光谱信号出现相邻谱峰重叠现象,给混合气体组成成分的定性或定量检测带来较大的困难。现有的方法在获取先验知识、处理精度、运算效率等方面存在不足。提出基于时频域分形维数分析的光谱信号重叠峰解析算法,结合小波的多尺度观测能力和分形的自相似度的度量能力,识别、定位和解析光谱信号中的重叠峰。首先利用小波对具有重叠谱峰的光谱信号进行光谱频率域和尺度域的分析,然后对该时频域的光谱信号在同一光谱频率下的多尺度数据进行自相似性度量和分形计算。逐频率计算后得到光谱信号在频率域的分形维数曲线。该曲线体现了光谱信号在不同尺度的自相似性,其极值位置与光谱信号的各独立峰的位置具有相关性。依据此特性,结合分形曲线的特征参数,最后利用神经网络解析出对应混合气体成分的混叠在一起的各个独立谱峰。该方法利用小波的多分辨率特性,对信号进行不同尺度的精细度量。分形模型则提高了系统解析复杂信号的能力,对重叠程度高的多谱峰重叠信号也有很强的处理能力。借助人工神经网络,实现了整个算法的自动测量。通过实验结果分析,验证了算法的有效性,并讨论影响算法效果的主要因素。     

基于不同散射机制特征的海杂波时变多普勒谱模型

海杂波的多普勒谱建模对采用多普勒处理技术的雷达进行有效的海杂波抑制和目标检测具有重要的意义.本文分别考虑Bragg,白冠和破碎波三种散射机制对应的多普勒谱分量的特征,对三种谱分量的频移和展宽进行分离,并引入附加速度频移量,提出了基于不同散射机制特征的雷达海杂波时变多普勒谱模型.该模型假设谱强度为受观测时间区间影响的随机变量,能够同时适用于平均多普勒谱与短时多普勒谱建模.通过分别对黄海海域实测的岸基P,S波段海杂波平均多普勒谱与短时多普勒谱建模测试,结果表明该模型相对传统模型的建模精度更高,尤其体现在短时谱的观测时间较长和平均谱形式较为复杂的情况下,建模误差显著减小.     

The time-singularity multifractal spectrum distribution

Although the multifractal singularity spectrum revealed the distribution of singularity exponent, it failed to consider the temporal information, therefore it is hard to describe the dynamic evolving process of non-stationary and nonlinear systems. In this paper, we aim for a multifractal analysis and propose a time-singularity multifractal spectrum distribution (TS-MFSD), which will hopefully reveal the spatial dynamic character of fractal systems. Similar to the Wigner–Ville time-frequency distribution, the time-delayed conjugation of fractal signals is selected as the windows function. Furthermore, the time-varying Holder exponent and the time-varying wavelet singularity exponent are deduced based on the instantaneous self-correlation fractal signal. The time-singularity exponent distribution i.e. TS-MFSD is proposed, which involves time-varying Hausdorff singularity spectrum distribution, time-varying large deviation multifractal spectrum and time-varying Legendre spectrum distribution, which exhibit the singularity exponent distribution of fractal signal at arbitrary time. Finally, we studied the algorithm of the TS-MFSD based on the wavelet transform module maxima method, analyzed and discussed the characteristic of TS-MFSD based on Devil Staircase signal, stochastic fractional motion and real sea clutter.     

The fractal energy measurement and the singularity energy spectrum analysis

The singularity exponent (SE) is the characteristic parameter of fractal and multifractal signals. Based on SE, the fractal dimension reflecting the global self-similar character, the instantaneous SE reflecting the local self-similar character, the multifractal spectrum (MFS) reflecting the distribution of SE, and the time-varying MFS reflecting pointwise multifractal spectrum were proposed. However, all the studies were based on the depiction of spatial or differentiability characters of fractal signals. Taking the SE as the independent dimension, this paper investigates the fractal energy measurement (FEM) and the singularity energy spectrum (SES) theory. Firstly, we study the energy measurement and the energy spectrum of a fractal signal in the singularity domain, propose the conception of FEM and SES of multifractal signals, and investigate the Hausdorff measure and the local direction angle of the fractal energy element. Then, we prove the compatibility between FEM and traditional energy, and point out that SES can be measured in the fractal space. Finally, we study the algorithm of SES under the condition of a continuous signal and a discrete signal, and give the approximation algorithm of the latter, and the estimations of FEM and SES of the Gaussian white noise, Fractal Brownian motion and the multifractal Brownian motion show the theoretical significance and application value of FEM and SES.     

Wavelet-based multifractal analysis of DNA sequences by using chaos-game representation

Chaos game representation (CGR) is proposed as a scale-independent representation for DNA sequences and provides information about the statistical distribution of oligonucleotides in a DNA sequence. CGR images of DNA sequences represent some kinds of fractal patterns, but the common multifractal analysis based on the box counting method cannot deal with CGR images perfectly. Here, the wavelet transform modulus maxima (WTMM) method is applied to the multifractal analysis of CGR images. The results show that the scale-invariance range of CGR edge images can be extended to three orders of magnitude, and complete singularity spectra can be calculated. Spectrum parameters such as the singularity spectrum span are extracted to describe the statistical character of DNA sequences. Compared with the singularity spectrum span, exon sequences with a minimal spectrum span have the most uniform fractal structure. Also, the singularity spectrum parameters are related to oligonucleotide length, sequence component and species, thereby providing a method of studying the length polymorphism of repeat oligonucleotides.     

Performance comparison of detection methods for weak target based on two-dimensional fractal sea surface

A detection method of the weak radar target is studied by applying fuzzy theory and multifractal correlation theory based on a two-dimensional fractal sea surface model. Firstly, a two-dimensional fractal sea surface model and its backscattering coefficient are introduced, the backscattering coefficient is a universal model affected by seawater permittivity, electromagnetic wave incidence angle, incident frequency, wind speed and wind direction factors. A novel two-dimensional wideband radar echo model, which is considered as a time-domain convolution of the stepped frequency signal radiated by airborne radar and the backscattering coefficient, is derived. Secondly, multifractal correlation theory is elaborated and a computation method of a membership degree of multifractal correlation spectrum is proposed, fuzzy theory and the AdaBoost algorithm are applied to the target detection. Finally, several target detection methods are compared with CA-CFAR and works of the predecessors. The results of the comparative study show its rationality of the two-dimensional wideband radar echo model and the superiority of wideband radars in detection performance, it is also seen that the multifractal correlation spectrum outperforms the multifractal spectrum in the probability of detection.     

Estimation of sea clutter inherent Doppler spectrum from shipborne S-band radar sea echo

Measurement of shipborne radar sea echo is instrumental in collecting the sea clutter data in open sea areas. However,the ship movement would introduce an extra Doppler component into the spectrum of the sea clutter, so the sea clutter inherent spectrum must be estimated prior to investigating the sea clutter Doppler characteristics from the shipborne radar sea echo. In this paper we show some results about a shipborne sea clutter measurement experiment that was conducted in the South China Sea in a period between 2017 and 2018; abundant clutter data have been collected by using a shipborne S-band clutter measurement radar. To obtain the sea clutter inherent Doppler spectrum from these data, an estimation method, based on the mapping relationship between the shipborne clutter spectrum and the inherent clutter spectrum, is proposed. This method is validated by shipborne clutter data sets under the same measuring conditions except for the ship speed. Using this method, the characteristics of the Doppler spectrum lineshapes in the South China Sea are calculated and analyzed according to different sea states, wave directions, and radar resolutions, which can be instrumental in designing the radar target detection algorithms.     

亚像素多重分形方法在图像处理中的应用

提出了一种用亚像素多重分形原理求取图像奇异性的新型算法,降低了单纯依靠整数像素位置灰度级梯度信息计算边缘测度所产生的误差。该算法结合CCD成像机理给出在亚像素位置的灰度级梯度分布规律,利用多重分形理论将实际图像分割成一系列具有不同奇异性指数的分形集合,对应着从边缘到纹理各层面的图像内容。模拟计算了投影小波中心点改变单位距离对边缘测度的影响程度,得出亚像素分割梯度的方法可以增加计算结果鲁棒性的结论。此方法用于标准图像的分割中,选用5×5亚像素数目提取的最奇异性集合与索贝尔(Sobel)算子(默认阈值为36.7920)提取的边缘的峰值信噪比为9.3981 dB。应用于复杂路面的裂纹提取中,其结果更符合人类的视觉观测。     

Sea Clutter Suppression and Target Detection Algorithm of Marine Radar Image Sequence Based on Spatio-Temporal Domain Joint Filtering

In marine radar target detection, sea clutter will cause a large number of missed alarms and false alarms, which will affect the accuracy of target detection. In order to suppress sea clutter effectively, a sea clutter suppression and target detection algorithm of marine radar image sequence based on spatio-temporal domain joint filtering is proposed in this paper. The proposed method is to add a sea clutter suppression link before detecting the target. Firstly, the marine radar image sequence is transformed into three-dimensional frequency wavenumber domain by three-dimensional fast Fourier transform (3D-FFT), and then the three-dimensional image spectrum is obtained. According to the fact that the sea clutter spectrum obtained from the image spectrum satisfies the dispersion relation of linear wave theory in the three-dimensional frequency wavenumber domain, a sea clutter model is established. Then, through the established sea clutter model, a spatio-temporal domain joint sea clutter suppressor is designed to filter the image spectrum. After that, the filtered image spectrum is transformed by three-dimensional inverse fast Fourier transform (3D-IFFT) to obtain the image sequence in which sea clutter is suppressed. Finally, target detection is carried out for sea clutter suppressed image sequence. The method is validated by using the real data of X-band marine radar. Compared with the classical Empirical mode decomposition (EMD) method, the improvement of signal-to-noise ratio (SNR) is more obvious, and SNR can be increased by 15.3 db at most. In addition, compared with target detection on original images directly, the proposed method has excellent detection rate and can increase detection rates by at least 8%.