基于多特征多核学习的全极化SAR溢油提取方法

全极化合成孔径雷达(Synthetic Aperture Radar,SAR)数据具有丰富的极化信息,能够提取出大量异构性特征。核学习方法在解决小样本、高维特征分类问题上具有优势,但异构特征对不同核函数具有响应差异。本文利用一种引入先验标签的多核学习方法进行全极化SAR的溢油信息提取,即基于分析结果对特征集进行遴选与组合,分别在每个特征组合中训练得到一个预备层核函数,以新获取的预备层核函数作为新的底层核函数,对全部特征进行学习分类。通过提取与分析溢油和海水的统计特征、物理散射特征和纹理特征,建立溢油全极化SAR特征谱,并利用引入先验标签的多核学习分类器进行溢油提取实验。结果表明,该方法能够利用全极化SAR多维异构特征的互补特性有效提高溢油分类提取精度。     

海上溢油极化特征及其探测研究

The SAR(Synthetic Aperture Radar) has the capabilities for all-weather day and night use. In the case of determining the effects of oil spill dumping, the oil spills areas are shown as dark spots in the SAR images.Therefore, using SAR data to detect oil spills is becoming progressively popular in operational monitoring, which is useful for oceanic environmental protection and hazard reduction. Research has been conducted on the polarization decomposition and scattering characteristics of oil spills from a scattering matrix using allpolarization of the SAR data, calculation of the polarization parameters, and utilization of the CPD(Co-polarized Phase Difference) of the oil and the sea, in order to extract the oil spill information. This method proves to be effective by combining polarization parameters with the characteristics of oil spill. The results show that when using Bragg, the oil spill backscattering machine with Enopy and a mean scatter α parameter. The oil spill can be successfully identified. However, the parameter mechanism of the oil spill remains unclear. The use of CPD can easily extract oil spill information from the ocean, and the polarization research provides a base for oil spill remote sensing detection.     

基于简缩极化SAR的溢油检测与分类方法

针对简缩极化SAR在海上溢油的检测与分类应用开展研究,利用欧式距离全面分析了简缩极化SAR的36种极化特征在溢油检测与油膜分类中的性能,发现简缩极化特征中的奇次散射系数的溢油检测性能最好,简缩极化熵的疑似溢油鉴别性能最好.在此基础上,提出了结合二叉树原理的简缩极化SAR溢油检测与油膜分类算法,并分析了RADARSAT-...     

基于GA-WNN的极化SAR海洋溢油检测方法研究

海洋溢油对海洋生态和人类生活带来严重的影响。由于合成孔径雷达(Synthetic Aperture Radar,SAR)具有全天时全天候的工作能力,在海洋溢油检测中发挥重要作用。目前,极化SAR是SAR探测技术的先进手段。本文利用6个极化特征进行溢油检测,通过对比分析这些特征对不同溢油的检测能力,得出单一极化特征在溢油检测中存在不足。通过J-M特征优选方法,提取出溢油检测识别度较高的特征影像,并利用遗传算法优化的小波神经网络(Genetic Algorithm-Wavelet Neural Network,GA-WNN)进行溢油检测。利用2套Radarsat-2全极化数据进行了方法验证,结果表明,该方法优于其他检测方法,溢油检测精度分别达到90.31%和95.42%。     

SAR极化特征图像与σ0图像的海洋内波可视性对比

利用ALOS PALSAR全极化SAR内波图像,对比分析了SAR海洋内波在11种极化特征与后向散射系数(σ⁰) 图像中的可视性。在提取的SAR极化特征图像中,Lambda值的内波特征最为明显,极化熵和极化角次之。与σ⁰图像相比,Lambda值的内波可视性优于同极化的σ⁰图像;对于极化熵和极化角,沿距离向传播的内波可视性优于同极化的σ⁰图像,沿方位向传播的内波可视性略差于同极化的σ⁰图像,两者均优于交叉极化的σ⁰图像。HH/VV极化比、归一化圆极化系数和Bata值的内波特征较弱;HH/HV极化比、VV/VH极化比、Gamma值、Delta值和各向异性指数的内波图像均不清晰,无法识别内波。     

结合极化分解特征的SVM溢油检测研究

Marine oil spills have caused major threats to marine environment over the past few years.The early detection of the oil spill is of great significance for the prevention and control of marine disasters.At present,remote sensing is one of the major approaches for monitoring the oil spill.Full polarization synthetic aperture radarc SAR data are employed to extract polarization decomposition parameters including entropy(H) and reflection entropy(A).The characteristic spectrum of the entropy and reflection entropy combination has analyzed and the polarization characteristic spectrum of the oil spill has developed to support remote sensing of the oil spill.The findings show that the information extracted from(1–A)×(1–H) and(1–H)×A parameters is relatively evident effects.The results of extraction of the oil spill information based on H×A parameter are relatively not good.The combination of the two has something to do with H and A values.In general,when H0.7,A value is relatively small.Here,the extraction of the oil spill information using(1–A)×(1–H) and(1–H)×A parameters obtains evident effects.Whichever combined parameter is adopted,oil well data would cause certain false alarm to the extraction of the oil spill information.In particular the false alarm of the extracted oil spill information based on(1–A)×(1–H) is relatively high,while the false alarm based on(1–A)×H and(1–H)×A parameters is relatively small,but an image noise is relatively big.The oil spill detection employing polarization characteristic spectrum support vector machine can effectively identify the oil spill information with more accuracy than that of the detection method based on single polarization feature.     

基于SAR极化比和纹理特征的海面溢油识别方法

针对海洋表面SAR影像的特点,采用基于灰度共生矩阵的纹理特征方法是提取海面溢油信息的常用方法,但实际海洋表面复杂的信息使得SAR图像上产生类似溢油现象的暗斑区域,这导致在利用纹理特征方法提取溢油信息时存在虚警率,降低了溢油信息的提取精度。基于RADARSAT-2 SAR四极化影像,本文提出基于SAR极化比影像的纹理特征识别方法对海面油膜进行识别提取。结果显示,基于SAR极化比影像的纹理特征识别方法可以有效且准确地提取海面溢油信息,相比于VV极化影像的纹理特征识别方法,溢油监测过程中的虚警率降低了17.96%,溢油监测总体精度达到96.83%。     

雷达波段对多极化SAR海面溢油检测极化特征参数的影响

多极化SAR数据海面溢油检测研究日益受到重视。本文研究不同波段极化SAR数据的海面溢油检测能力,为最大程度减小观测条件、环境因素等的影响,选取准同步获取的SIR-C/X多极化SAR数据。针对海面油膜、生物油膜和低风区疑似溢油现象,研究L波段和C波段的共极化相位差、一致性系数、极化熵、各向异性和平均散射角等极化特征对海面油膜以及不同海面暗斑现象的检测能力。研究结果表明:在海面溢油检测以及探测不同暗斑现象间差异方面,C波段总体优于L波段;L波段,极化分解特征各向异性参数优于共极化相位差和一致性系数;C波段,共极化相位差、一致性系数特征优于极化分解特征各向异性和极化熵,结合平均散射角特征有助于滤除生物油膜和低风区。     

基于极化参数的SVM海上溢油识别

溢油对海洋环境造成的危害越来越大,及早发现对于减灾防灾具有重要意义。目前,运用极化SAR进行溢油探测已成为遥感监测的一个重要方面,本文基于SIR-C数据,开展极化SAR的溢油监测,提取极化参数熵H,散射角α和反熵A,运用SVM监督分类方法,进行溢油信息提取。结果表明,基于SVM的分类精度要强于基于H-α分类的分类结果。     

基于改进Faster R-CNN模型的SAR图像溢油检测方法

SAR（synthetic aperture radar）图像溢油暗斑准确识别对海上溢油应急工作具有重要的意义。为减少SAR图像特征提取、特征选择过程中人为因素对溢油检测精度的影响,本文将Faster R-CNN卷积神经网络模型引入SAR图像溢油检测并进行了改进。针对溢油暗斑形状多样及SAR图像背景复杂的特点,选用结构一致且实用性强的VGG16卷积网络获取图像特征,并使用软化非极大值抑制算法（Soft-NMS）进行优化。同时基于相同的数据集,提取常用的SAR图像几何特征、灰度特征和纹理特征,构建反向传播（backpropagation,BP）人工神经网络溢油检测方法并与Faster R-CNN方法进行对比。实验结果表明,基于改进Faster-RCNN模型的溢油检测方法溢油检测率达到0.78,且溢油检测虚警率低于0.25,相比BP人工神经网络溢油检测方法样本识别率、溢油检测率分别提高了4%和5%,溢油虚警率降低了5%。     

全极化合成孔径雷达近岸风场反演研究

Coastal winds are strongly influenced by topology and discontinuity between land and sea surfaces. Wind assessment from remote sensing in such a complex area remains a challenge. Space-borne scatterometer does not provide any information about the coastal wind field, as the coarse spatial resolution hampers the radar backscattering. Synthetic aperture radar （SAR） with a high spatial resolution and all-weather observation abilities has become one of the most important tools for ocean wind retrieval, especially in the coastal area. Conventional methods of wind field retrieval from SAR, however, require wind direction as initial information, such as the wind direction from numerical weather prediction models （NWP）, which may not match the time of SAR image acquiring. Fortunately, the polarimetric observations of SAR enable independent wind retrieval from SAR images alone. In order to accurately measure coastal wind fields, this paper proposes a new method of using co-polarization backscattering coefficients from polarimetric SAR observations up to polarimetric correlation backscattering coefficients, which are acquired from the conjugate product of co-polarization backscatter and cross-polarization backscatter. Co-polarization backscattering coefficients and polarimetric correlation backscattering coefficients are obtained form Radarsat-2 single-look complex （SLC） data.The maximum likelihood estimation is used to gain the initial results followed by the coarse spatial filtering and fine spatial filtering. Wind direction accuracy of the final inversion results is 10.67 with a wind speed accuracy of 0.32 m/s. Unlike previous methods, the methods described in this article utilize the SAR data itself to obtain the wind vectors and do not need external wind directional information. High spatial resolution and high accuracy are the most important features of the method described herein since the use of full polarimetric observations contains more information about the space measured.This article is a useful addition to the work of independent SAR wind retrieval. The experimental results herein show that it is feasible to employ the co-polarimetric backscattering coefficients and the polarimetric correlation backscattering coefficients for coastal wind field retrieval.     

海洋溢油合成孔径雷达图像特征提取及其关键度分析

就SAR图像溢油检测的方法论而言,用于识别溢油和疑似现象的定性或定量的统计特征量选择,通常是任意的。对于不同的分类模型,所选用的特征量也不尽相同。主要是进行海洋SAR图像特征提取及其关键度分析。其目的是将"最小距离"判别法应用于海上溢油和疑似溢油的识别研究。首先,针对海洋SAR图像溢油检测常用的特征量,进行冗余处理;然后,引入关键系数,定量地研究特征量的关键度,提取显著特征量;藉以构造一个多维的特征矢量空间,以适于最小距离判别法在特征矢量空间中进行溢油和疑似溢油的识别研究。     

一种基于经验模态分解的SAR图像溢油识别算法

溢油事件的发生会给海洋环境的保护和经济发展带来巨大的影响。运用现代化的监测手段和技术进行监测,及时发现溢油现象和违规行为,保护海洋环境是非常重要的。合成孔径雷达（SAR）技术是溢油检测的有效工具,在SAR图像中溢油表现为黑色的区域,但是黑色区域也可能会由其他的因素引起。本文提出了一种基于二维经验模态分解（BEMD）的方法来识别溢油和疑似溢油。首先通过BEMD方法将感兴趣的区域分解为局部窄带的各分量—内蕴模函数（BIMF）之和,并对分解后得到的各分量IMF进行Hilbert变换,通过Hibert谱分析得到64维的特征空间,然后使用Relief方法得到5个特征向量,最后利用马氏距离分类器进行分类。通过实验结果表明,该方法能够有效、准确地检测出溢油,准确率超过90%。     

Semisupervised heterogeneous ensemble for ship target discrimination in synthetic aperture radar images

Ship detection using synthetic aperture radar (SAR) plays an important role in marine applications. The existing methods are capable of quickly obtaining many candidate targets, but numerous non-ship objects may be wrongly detected in complex backgrounds. These non-ship false alarms can be excluded by training discriminators, and the desired accuracy is obtained with enough verified samples. However, the reliable verification of targets in large-scene SAR images still inevitably requires manual interpretation, which is difficult and time consuming. To address this issue, a semisupervised heterogeneous ensemble ship target discrimination method based on a tri-training scheme is proposed to take advantage of the plentiful candidate targets. Specifically, various features commonly used in SAR image target discrimination are extracted, and several acknowledged classification models and their classic variants are investigated. Multiple discriminators are constructed by dividing these features into different groups and pairing them with each model. Then, the performance of all the discriminators is tested, and better discriminators are selected for implementing the semisupervised training process. These strategies enhance the diversity and reliability of the discriminators, and their heterogeneous ensemble makes more correct judgments on candidate targets, which facilitates further positive training. Experimental results demonstrate that the proposed method outperforms traditional tri-training.     

台风中降雨对中国高分3号合成孔径雷达观测信号的影响

Gaofen-3(GF-3), a Chinese civil synthetic aperture radar(SAR) at C-band, has operated since August 2016.Remarkably, several typhoons have been captured by GF-3 around the China Seas over its last two-year mission.In this study, six images acquired in Global Observation(GLO) and Wide ScanSAR(WSC) modes at verticalvertical(VV) polarization channel are discussed. This work focuses on investigating the observation of rainfall using GF-3 SAR. These images were collocated with winds from the European Centre for Medium-Range Weather Forecasts(ECMWF), significant wave height simulated from the WAVEWATCH-III(WW3) model, sea surface currents from climate forecast system version 2(CFSv2) of the National Centers for Environmental Prediction(NCEP) and rain rate data from the Tropical Rainfall Measuring Mission(TRMM) satellite. Sea surface roughness,was compared with the normalized radar cross section(NRCS) from SAR observations, and indicated a 0.8 correlation(COR). We analyzed the dependences of the difference between model-simulated NRCS and SARmeasured NRCS on the TRMM rain rate and WW3-simulated significant wave height. It was found that the effects of rain on SAR damps the radar signal at incidence angles ranging from 15° to 30°, while it enhances the radar signal at incidence angles ranging from 30° to 45° and incidence angles smaller than 10°. This behavior is consistent with previous studies and an algorithm for rain rate retrieval is anticipated for GF-3 SAR.     

复合雷达后向散射模型与合成孔径雷达、散射计和高度计海面雷达后向散射观测的比较分析

海洋微波散射模型相比于以经验统计建立的地球物理模式函数具有不受特定微波频率限制的优势。组合布拉格散射模型和几何光学模型形成了复合雷达后向散射模型。利用南海北部气象浮标2014年海面风速风向实测值作为散射模型输入,分别比较了复合雷达后向散射模型与RADARSAT-2卫星C波段SAR、HY-2A卫星Ku波段微波散射计的海面后向散射系数,偏差分别为(?0.22±1.88) dB (SAR)、(0.33±2.71) dB (散射计VV极化）和(?1.35±2.88) dB (散射计HH极化);以美国浮标数据中心(NDBC)浮标2011年10月1日至2014年9月30日共3年的海面风速、风向实测值作为散射模型输入,分别比较了复合雷达后向散射模型与Jason-2、HY-2A卫星Ku波段高度计海面后向散射系数,偏差分别为(1.01±1.15) dB和(1.12±1.29) dB。中等入射角和垂直入射下的卫星传感器后向散射系数观测值与复合雷达后向散射模型模拟值比较,具有不同的偏差,但具有相同的海面风速检验精度,均方根误差小于1.71 m/s。结果表明,复合雷达后向散射模型可模拟计算星载SAR、散射计和高度计观测条件下的海面雷达后向散射系数,且与CMOD5、NSCAT-2、高度计业务化海面风速反演的地球物理模式函数的计算结果具有一致性;复合雷达后向散射模型可用于微波遥感器的定标与检验、海面雷达后向散射的模拟。     

一种基于Curvelet变换的合成孔径雷达海洋内波图像的斑点噪声抑制方法

合成孔径雷达是海洋内波研究中最重要的工具之一。雷达图像中的斑点噪声会严重降低图像的质量,这一问题在处理和分析信号较弱的二模态内波信号和上升型内波信号时极为明显。合成孔径雷达图像中的海洋内孤立波的信号具有明显的尺度性和方向性。同时,curvelet变换作为一种同时具备尺度分辨率和方向分辨率的数学变换,能够对一幅雷达图像在不同尺度、不同方向和不同位置上进行分析。本文给出了一个基于curvelet变换的合成孔径雷达海洋内孤立波图像的斑点噪声抑制方法。该方法可简述为：(1)对一幅合成孔径雷达海洋内孤立波图像进行curvelet变换,获得curvelet系数;(2)分别仅仅保留一个尺度的系数,将其它尺度的系数置为零,利用处理之后的系数分别重建图像,得到仅仅用一个尺度的系数重建的图像;(3)分别计算上一步中得到的图像的均方差,根据波浪理论,图像的方差代表能量,方差越大则能量越大,以此可以确定内波信息集中的尺度;(4)在每个尺度下,分别计算每个方向的curvelet系数矩阵的平均值,以此确定内孤立波信号集中的方向;(5)在上两步工作的基础上,仅仅保留内波信号集中的尺度和方向的系数,而将其它尺度和方向的系数置为0,得到一幅提取主波信息的图像;(6)将上一步得到的提取主波信息的图像加回到原始图像中,从而达到增强波浪信息并抑制斑点噪声的目的。大量的实验验证表明,该方法不仅能有效地压制斑点噪声,而且能有效地增强波浪信号。     

极化X波段雷达海面溢油微波散射模型研究

为研究溢油对海面电磁散射的影响,作者根据海面复合微波散射模型理论和蒙特卡洛统计模型理论,通过引入单分子油膜的黏性阻尼,对粗糙的溢油海面进行建模,定量分析溢油对海浪谱和海面后向散射系数值两个方面的影响。为实现基于X波段雷达海面溢油检测提供理论支撑,有助于解决溢油检测中的虚警率高的问题。     

基于影像纹理特征和外部风向的星载 SAR 海面风场反演研究

在合成孔径雷达(Synthetic Aperture Radar,SAR)海面风场反演中,基于风条纹影像纹理特征的海面风向反演方法精度高,但是依赖于图像风条纹的存在,而外部风向信息与SAR资料时空分辨率不易匹配、精度较低,从而影响大面积、高分辨率海面风场反演的精度。针对此问题,提出一种将SAR图像风条纹线性纹理特征与外部风向信息相结合的星载SAR海面风向获取方法,在SAR影像线性纹理特征明显的区域采用二维连续小波变换得到高精度的海面风向,其余区域采用与之时空相匹配的数值预报模式风向填充;并利用地球物理模型函数进一步得到海面风速,进而实现高精度、大范围海面风场的反演。为验证本文方法的有效性,利用ENVISAT/ASAR数据进行风场反演试验,并将反演结果与浮标实测数据进行比对。结果表明:在线性纹理特征明显的区域,小波方法的反演精度优于快速傅里叶变换(FFT)法和数值预报模式风向;外部风向精度略低,但与SAR观测资料时空匹配性较好,弥补了风条纹风向的不足。二者的结合为星载SAR海面风场反演的业务化应用提供了支持。