基于改进二维Ostu算法的SAR图像分割方法研究

INS/SAR组合导航系统对SAR图像分割提出了较高的实时性和准确性要求。文章针对SAR图像的特点与二维Ostu方法在区域划分以及运行速度上的缺陷,提出了一种改进的二维Ostu方法,更为准确地划分了目标和背景区域,并且将阈值搜索空间由二维降为一维。经仿真验证,该方法具有良好的分割效果并能有效提高分割速度,满足了组合导航系统相应的设计要求。     

基于改进的二维Otsu法的图像分割法

针对Otsu阈值法仅适合图像目标和背景分布呈正态分布且各像素量和方差基本相当的不足,提出了一种基于二维Otsu的改进算法,实验结果表明,该算法比传统Otsu阈值法能够获得较好的分割效果.     

一种基于MLS的SAR图像点特征提取方法

针对合成孔径雷达图像点特征提取问题,提出了一种基于MLS的SAR图像点特征提取方法.首先利用提出的基于超像素关联分析的SAR图像分割算法得到二值分割图像,再把该二值图像和SAR图像进行点乘运算,获取到含有强度信息的目标区域,然后采用移动最小二乘法(MLS)对目标区域进行曲面拟合,根据设定的点特征判决规则,最后提取出SAR图像的峰值、脊和谷等多种点特征.基于MSTAR数据的实验结果表明了该方法的有效性和准确性.     

利用α稳定分布的小波域SAR图像降斑算法

提出了一种改进的合成孔径雷达(SAR)图像α-MAP降斑算法,通过采用厚尾α稳定分布和引入噪声方差调整系数的方法,可以获得更好的降斑和纹理保持效果,同时应用平稳小波分解来解决Gibbs效应引入的伪噪声问题。最后通过对实际SAR图像的降斑处理,对本文的算法与最新的小波软阈值算法以及传统的局部自适应滤波器在性能参数和滤波效果上做了全面的比较。试验表明本文的算法更具适应性,实现了降斑与纹理保持更有效的综合滤波,在性能参数和复原图像效果上都有较好的表现。     

基于Canny的改进图像边缘检测算法

边缘检测是图像处理领域中最重要的关键技术之一。针对经典边缘检测算法抗椒盐噪声性能较差及阈值选取适应性不强等问题,提出了一种基于Canny的算法架构,结合自适应中值滤波(Adaptive Median Filtering,AMF）、大津法(Otsu）以及最大熵法(Maximum Entropy Method,MEM）的改进图像边缘检测算法。该算法首先结合改进自适应中值滤波对图像降噪,从而在保留图像细节的同时较好地滤除了椒盐噪声干扰。而后利用基于Otsu和MEM提出的改进双阈值选取方法,获取自适应的高低阈值对图像边缘进行检测,边缘检测准确度可以达到96%以上。实验结果表明,本文算法在椒盐噪声干扰下针对背景复杂的图像有更好的边缘检测效果。     

超燃冲压发动机燃烧流场纹影图像无监督分割方法

超燃冲压发动机燃烧室流场纹影图像常存在大量噪声信号,如何高效、准确提取燃烧流场图像的主要波系结构成为当前亟需探索的问题。以超燃冲压发动机燃烧室内冷流到氢燃料点火阶段流场为研究对象,基于深度神经网络方法,发展一种燃烧室内流场的关键波系结构快速识别方法。首先,采用基于图论的超像素分割方法对纹影图像进行聚类分割,为语义信息明显相同区域分配伪标签;其次,设计了一种全卷积特征提取神经网络,并使用残差结构对各个通道进行加权,提取纹影图像高级语义特征;最后,使用交叉熵目标函数优化网络模型,并通过阈值滤波操作去除噪声像素点,提升语义分割效果。结果表明：与K-means及自适应高斯阈值方法相比,本文提出方法在准确率、召回率、F1分数和交并比指标性能明显是最优的,能够准确完成燃烧流场纹影图像语义分割任务,可以更加清晰地反应流场内的主要波系和剪切层结构     

自适应K-means图像分割方法

本文提出了一种自适应K-means图像分割方法,该方法通过用Otsu方法来初步选取阈值作为初始K-means聚类中心,然后经过K-means聚类方法对图像进行分割,最后根据K-means获得的聚类中心与Otsu所确定的阈值进行平均计算,作为图像分割的阈值.实验结果表明:该方法自动快速形成的分割阈值较为合理,对图像能达到更好的分割效果,可以进一步提高后续图像匹配与定位的准确性.     

基于二维经验小波变换的金属断口图像消噪方法研究

针对传统小波变换在金属断口图像分析中存在的不足,结合二维经验模态分解和小波分析,本文提出了一种基于二维经验小波的断口图像消噪算法。该方法通过自适应构造滤波器组构造了经验小波函数,实现了在频域上分割傅里叶频谱的同时又分离出了信号的不同模式,从而提取了具有紧支撑傅立叶频谱的AM-FM成分。将所提方法与传统小波变换方法进行对比,结果表明,不论是从峰值信噪比还是从结构相似指数来看,二维经验小波算法的图像消噪效果都明显优于小波变换方法。最后,将二维经验小波算法应用到金属断口图像消噪中,进一步验证了所提方法的有效性。     

基于序列二次规划优化阈值的NSCT高斯噪声图像滤波方法

提出了一种基于序列二次规划(SQP)优化阈值的非下采样Contourlet变换(NSCT)图像高斯白噪声去除方法。该方法利用广义交叉验证(GCV)准则作为优化指标,使用序列二次规划算法对NSCT域的去噪阈值进行优化,能够在噪声方差等图像先验知识未知的情况下得到最优去噪阈值。确定阈值后,采用非线性阈值函数对Contourlet系数进行处理。实验结果表明与其他Contourlet域去噪方法相比,该方法能有效去除图像的高斯白噪声,提高图像的峰值信噪比,并较好地保留图像的边缘信息。     

基于小波域贝叶斯估计模糊萎缩的SAR图像降斑算法

提出了基于小波域高斯混合模型贝叶斯估计模糊萎缩的SAR图像降斑算法.该算法分析了SAR图像在平稳小波变换(SWT)域中的统计模型,并用高斯混合模型对其进行描述,推导出基于贝叶斯估计的信号最小均方误差(MMSE)的模糊萎缩因子.籍此再根据小波域相邻尺度间小波系数的相关性,采用分区域模糊萎缩思想,很好地得到无斑点真实信号小波系数的估计值.仿真结果表明该算法在大大抑制斑点噪声的同时,有效地保持了边缘,其性能优于改进Lee滤波、小波软阈值和SWT萎缩降斑算法.     

High-resolution SAR imaging with angular diversity

We propose to use the APES (amplitude and phase estimation) approach for the spectral estimation of gapped data and synthetic aperture radar (SAR) imaging with angular diversity. A relaxation-based algorithm, referred to as GAPES (Gapped-data APES), is proposed, which includes estimating the spectrum via APES and filling in the gaps via a least squares (LS) fitting. For SAR imaging with angular diversity data fusion, we perform one-dimensional (1-D) windowed fast Fourier transforms (FFTs) in range, use the GAPES algorithm to interpolate the gaps in the aperture for each range, apply 1-D inverse FFTs (IFFTs) and dewindow in range, and finally apply the two-dimensional (2-D) APES algorithm to the interpolated matrix to obtain the 2-D SAR image. Numerical results are presented to demonstrate the effectiveness of the proposed algorithm     

SAR image formation via semiparametric spectral estimation

A new algorithm, referred to as the SPAR (Semiparametric) algorithm, is presented herein for target feature extraction and complex image formation via synthetic aperture radar (SAR). The algorithm is based on a flexible data model that models each target scatterer as a two-dimensional (2-D) complex sinusoid with arbitrary unknown amplitude and constant phase in cross-range and with constant amplitude and phase in range. By attempting to deal with one corner reflector, such as one dihedral or trihedral, at a time, the algorithm can be used to effectively mitigate the artifacts in the SAR images due to the flexible data model. Another advantage of SPAR is that it can be used to obtain initial conditions needed by other parametric target feature extraction methods to reduce the total amount of computations needed. Both numerical and experimental examples are provided to demonstrate the performance of the proposed algorithm     

OGSRN: Optical-guided super-resolution network for SAR image

Although Convolutional Neural Networks(CNNs) have significantly improved the development of image Super-Resolution(SR) technology in recent years, the existing SR methods for SAR image with large scale factors have rarely been studied due to technical difficulty. A more efficient method is to obtain comprehensive information to guide the SAR image reconstruction.Indeed, the co-registered High-Resolution(HR) optical image has been successfully applied to enhance the quality of SAR image due to it...     

改进的自适应阈值软硬折中Curvelet图像去噪技术

Curvelet变换是继小波变换和Ridgelet变换之后,更适合图像处理的一种多尺度变换,它能同时获得对图像平滑区域和边缘部分的稀疏表达,且具有很强的方向性。针对软阀值和硬阈值去噪方法存在的不足,提出了基于Curvelet变换域的软硬阈值折中去噪法,并采用不同的阈值自适应地对不同的Curvelet子带进行阈值化,实验结果表明该方法对图像中的边缘曲线特征有更好的复原。去噪后图像PSNR值更高,视觉效果更好。     

Target prescreening based on a quadratic gamma discriminator

This work presents the development, analysis and validation of a new target discrimination module for synthetic aperture radar (SAR) imagery based on an extension of gamma functions to 2-D. Using the two parameter constant false-alarm rate (CFAR) stencil as a prototype, a new stencil based on 2-D gamma functions is used to estimate the intensity of the pixel under test and its surroundings. A quadratic discriminant function is created from these estimates, which is optimally adapted with least squares in a training set of representative clutter and target chips. This discriminator is called the quadratic gamma discriminator (QGD). The combination of the CFAR and the QGD was tested in realistic SAR environments and the results show a large improvement of the false alarm rate with respect to the two-parameter CFAR, both with high resolution (1 ft) fully polarimetric SAR and with one polarization, 1 m SAR data     

运动补偿用惯性器件误差对SAR成像分辨率的影响研究

 为提高合成孔径雷达（SAR）系统的性价比,必须根据SAR成像分辨率的要求和整体系统参数,设计相应精度的运动补偿用捷联惯导系统。在确定SAR运动补偿系统方案和安装方式的基础上,分析不同方向的加速度计和陀螺仪误差对天线相位中心位置测量误差的影响,并利用位置测量误差与SAR成像分辨率之间的关系,进一步明确了不同方向的加速度计和陀螺仪对SAR成像分辨率的影响。研究表明：基于SAR的工作原理和安装方式,x方向加速度计和y方向陀螺仪对SAR成像分辨率的影响明显比其他惯性器件严重;相同误差水平的惯性器件对SAR成像分辨率的影响随着合成孔径时间和工作波长的不同而不同,时间越长,波长越短,影响则越严重。SAR成像仿真证明了结论的正确性。研究结果对于研制高性价比SAR成像运动补偿系统有一定的理论指导意义。     

基于小波变换的图像低通去噪

提出了一种利用小波变换的多分辨率特性对图像进行低通去噪的方法,系统的阐述了其基本原理和核心算法,并对加噪后的图像实现了低通去噪处理。     

Super resolution SAR imaging via parametric spectral estimationmethods

Super resolution synthetic aperture radar (SAR) image formation via sophisticated parametric spectral estimation algorithms is considered. Parametric spectral estimation methods are devised based on parametric data models and are used to estimate the model parameters. Since SAR images rather than model parameters are often used in SAR applications, we use the parameter estimates obtained with the parametric methods to simulate data matrices of large dimensions and then use the fast Fourier transform (FFT) methods on them to generate SAR images with super resolution. Experimental examples using the MSTAR and Environmental Research Institute of Michigan (ERIM) data illustrate that robust spectral estimation algorithms can generate SAR images of higher resolution than the conventional FFT methods and enhance the dominant target features     

On the resolvability of sinusoidal parameter estimates [andapplication to SAR target features]

The resolvability of 2-D (two-dimensional) sinusoidal parameter estimates is studied. These sinusoids describe the target features in SAR (synthetic aperture radar) applications. We analyze the resolvability by considering the frequency estimates of the sinusoids. Our results may be used by target classification algorithms to better classify radar targets in SAR applications