MULTI—SOURCE REMOTE SENSING IMAGE FUSION BASED ON SUPPORT VECTOR MACHINE

Remote Sensing image fusion is an effective way to use the large volume of data from multi-source images.This paper introduces a new method of remote sensing image fusion based on support vector machine(SVM),using high spatial resolution data SPIN-2 and multi-spectral remote sensing data SPOT-4.Firstly,the new method is established by building a model of remote sensing image fusion based on SVM.Then by using SPIN-2 data and SPOT-4 data ,image classify-cation fusion in tested.Finally,and evaluation of the fusion result is made in two ways.1)From subjectivity assessment,the spatial resolution of the fused image is improved compared to the SPOT-4.And it is clearly that the texture of the fused image is distinctive.2)From quantitative analysis,the effect of classification fusion is better.As a whole ,the re-sult shows that the accuracy of image fusion based on SVM is high and the SVM algorithm can be recommended for applica-tion in remote sensing image fusion processes.     

An Approach to Underwater Image Enhancement Based on Image Structural Decomposition

Underwater imaging posts a challenge due to the degradation by the absorption and scattering occurred during light propagation as well as poor lighting conditions in water medium. Although image filtering techniques are utilized to improve image quality effectively, problems of the distortion of image details and the bias of color correction still exist in output images due to the complexity of image texture distribution. This paper proposes a new underwater image enhancement method based on image structural decomposition. By introducing a curvature factor into the Mumford_Shah_G decomposition algorithm, image details and structure components are better preserved without the gradient effect. Thus, histogram equalization and Retinex algorithms are applied in the decomposed structure component for global image enhancement and non-uniform brightness correction for gray level and the color images, then the optical absorption spectrum in water medium is incorporate to improve the color correction. Finally, the enhanced structure and preserved detail component are recomposed to generate the output. Experiments with real underwater images verify the image improvement by the proposed method in image contrast, brightness and color fidelity.     

SAR image coregistration using fringe definition detection

In order to overcome the limitation of cross correlation coregistration method for Synthetic Aperture Radar(SAR) interferometric pairs with low coherence,a new image coregistration algorithm based on Fringe Definition Detection(FDD) is presented in this paper.The Fourier transformation was utilized to obtain spectrum characteristics of interferometric fringes.The ratio between spectrum mean and peak was proposed as the evaluation index for identifying homologous pixels from interferometric images.The satellites ERS-1/2 C-band SAR acquisitions covering the Yangtze River plain delta,eastern China and ALOS/PALSAR L-band images over the Longmen Shan mountainous area,southwestern China were respectively employed in the experiment to validate the proposed coregistration method.The testing results suggested that the derived Digital Elevation Model(DEM) from FDD method had good agreement with that from the cross correlation method as well as the reference DEM at high coherence area.However,The FDD method achieved a totally improved topographic mapping accuracy by 24 percent in comparison to the cross correlation method.The FDD method also showed better robustness and achieved relatively higher performance for SAR image coregistration in mountainous areas with low coherence.     

INTERNAL TIDES, SOLITARY WAVES AND BORES IN SHALLOW SEAS

Remote sensing and in situ observations of internal tides, solitary waves and bores in shallow water are briefly reviewed in this paper. The emphasis is laid on interpreting SAR images based on oceanographic measurements, and analyzing characteristics of internal waves in the China Seas. Direc-tions for future research are discussed.     

Underwater Target Detection Based on Reinforcement Learning and Ant Colony Optimization

Underwater optical imaging produces images with high resolution and abundant information and hence has outstanding advantages in short-distance underwater target detection.However,low-light and high-noise scenarios pose great challenges in un-derwater image and video analyses.To improve the accuracy and anti-noise performance of underwater target image edge detection,an underwater target edge detection method based on ant colony optimization and reinforcement learning is proposed in this paper.First,the reinforcement learning concept is integrated into artificial ants’movements,and a variable radius sensing strategy is pro-posed to calculate the transition probability of each pixel.These methods aim to avoid undetection and misdetection of some pixels in image edges.Second,a double-population ant colony strategy is proposed,where the search process takes into account global search and local search abilities.Experimental results show that the algorithm can effectively extract the contour information of underwater targets and keep the image texture well and also has ideal anti-interference performance.     

Edge detection of potential field data based on image processing methods

The conventional methods of edge detection can roughly delineate edge position of geological bodies, but there are still some problems such as low detection accuracy and being susceptible to noise interference.In this paper, three image processing methods, Canny, LoG and Sobel operators are briefly introduced, and ap-plied to edge detection to determine the edge of geological bodies.Furthermore, model data is built to analyze the edge detection ability of this image processing methods, and compare with conventional methods.Combined with gravity anomaly of Sichuan basin and magnetic anomaly of Zhurihe area, the detection effect of image pro-cessing methods is further verified in real data.The results show that image processing methods can be applied to effectively identify the edge of geological bodies.Moreover, when both positive and negative anomalies exist and noise is abundant, fake edge can be avoided and edge division is clearer, and satisfactory results of edge detection are obtained.     

Object-oriented crop classification based on UAV remote sensing imagery

UAV remote sensing images have the advantages of high spatial resolution,fast speed,strong real-time performance,and convenient operation,etc.,and have become a recently developed,vital means of acquiring surface information.It is an important research task for precision agriculture to make full use of the spectrum,texture,color and other characteristic information of crops,especially the spatial arrangement and structure information of features,to explore effective methods for the classification of multiple varieties of crops.In order to explore the applicability of the object-oriented method to achieve accurate classification of UAV high-resolution images,the paper used the object-oriented classification method in ENVI to classify the UAV high-resolution remote sensing image obtained from the orderly structured 28 species of crops in the test field,which mainly includes image segmentation and object classification.The results showed that the plots obtained after classification were continuous and complete,basically in line with the actual situation,and the overall accuracy of crop classification was 91.73%,with Kappa coefficient of 0.87.Compared with the crop planting area based on remote sensing interpretation and field survey,the area error of 17 species of crops in this study was controlled within 15%,which provides a basis for object-oriented crop classification of UAV remote sensing images.     

Auto-registration and orthorecification algorithm for the time series HJ-1A/B CCD images

How to deal with geometric distortion is an open problem when using the massive amount of satellite images at a national or global scale,especially for multi-temporal image analysis.In this paper,an algorithm is proposed to automatically rectify the geometric distortion of time-series CCD multispectral data of small constellation for environmental and disaster mitigation(HJ-1A/B) which was launched by China in 2008.In this algorithm,the area-based matching method was used to automatically search tie points firstly,and then the polynomial function was introduced to correct the systematic errors caused by the satellite motion along the roll,pitch and yaw direction.The improved orthorectification method was finally used to correct pixel displacement caused by off-nadir viewing of topography,which are random errors in the images and cannot be corrected by the polynomial equation.Nine scenes of level 2 HJ CCD images from one path/row were taken as the warp images to test the algorithm.The test result showed that the overall accuracy of the proposed algorithm was within 2 pixels(the average residuals were 37.8 m,and standard deviations were 19.8 m).The accuracies of 45.96% validation points(VPs) were within 1 pixel and 90.33% VPs were within 2 pixels.The discussion showed that three main factors including the distortion patterns of HJ CCD images,percent of cloud cover and the varying altitude of the satellite orbit may affect the search of tie points and the accuracy of results.Although the influence of varying altitude of the satellite orbits is less than the other factors,it is noted that detailed satellite altitude information should be given in the future to get a more precise result.The proposed algorithm should be an efficient tool for the geo-correction of HJ CCD multi-spectral images.     

Fusion and classification of Beijing-1 small satellite remote sensing image for land cover monitoring in mining area

In order to promote the application of Beijing-1 small satellite(BJ-1) remote sensing data,the multispectral and panchromatic images captured by BJ-1 were used for land cover classification in Pangzhuang Coal Mining.An improved Intensity-Hue-Saturation(IHS) fusion algorithm is proposed to fuse panchromatic and multispectral images,in which intensity component and panchromatic image are combined using the weights determined by edge pixels in the panchromatic image identified by grey absolute correlation degree.This improved IHS fusion algorithm outper-forms traditional IHS fusion method to a certain extent,evidenced by its ability in preserving spectral information and enhancing spatial details.Dempster-Shafer(D-S) evidence theory was adopted to combine the outputs of three member classifiers to generate the final classification map with higher accuracy than that by any individual classifier.Based on this study,we conclude that Beijing-1 small satellite remote sensing images are useful to monitor and analyze land cover change and ecological environment degradation in mining areas,and the proposed fusion algorithms at data and decision levels can integrate the advantages of multi-resolution images and multiple classifiers,improve the overall accuracy and produce a more reliable land cover map.     

Underwater topography detection of Taiwan Shoal with SAR images

Under suitable conditions of tidal current and wind, underwater topography can be detected by synthetic aperture radar (SAR) indirectly. Underwater topography SAR imaging includes three physical processes: radar ocean surface backscattering, the modulation of sea surface short wave spectrum by the variations in sea surface currents, and the modulation of sea surface currents by the underwater topography. The first process is described usually by Bragg scattering theory because the incident angle of SAR is always between 20°-70°. The second process is described by the action balance equation. The third process is described by an ocean hydrodynamic model. Based on the SAR imaging mechanism for underwater topography, an underwater topography SAR detection model and a simplified method for its calculation are introduced. In the detection model, a two-dimensional hydrodynamic model – the shallow water model is used to describe the motion of tidal current. Due to the difficulty of determining the expression of SAR backscattering cross section in which some terms can not be determined, the backscattering cross section of SAR image used in the underwater topography SAR detection is pro-processed by the simulated SAR image of the coarse-grid water depth to simplify the calculation. Taiwan Shoal, located at the southwest outlet of Taiwan Strait, is selected as an evaluation area for this technique due to the occurrence of hundreds of sand waves. The underwater topography of Taiwan Shoal was detected by two scenes of ERS-2 SAR images which were acquired on 9 January 2000 and 6 June 2004. The detection results are compared with in situ measured water depths for three profiles. The average absolute and relative errors of the best detection result are 2.23 m and 7.5 %, respectively. These show that the detection model and the simplified method introduced in the paper is feasible.     

Preformance evaluation of several fusion approaches for CCD/SAR images

Several image fusion approaches for CCD/SAR images are studied and the performance evaluation of these fusion approaches is completed in this paper. Firstly, the preprocessing of CCD/SAR images before fusion is fulfilled. Then, the image fusion methods including linear superposition, nonlinear operator method and multiresolution methods, of which the multiresolution methods include Laplacian pyramid, ratio pyramid, contrast pyramid, gradient pyramid, morphological pyramid and discrete wavelet transform, are adopted to fuse two types of images. Lastly, the four performance measures, standard deviation, entropy, cross entropy and spatial frequency, are calculated to compare the fusion results by different fusion approaches in this paper. Experimental results show that contrast pyramid, morphology pyramid and discrete wavelet transformation in multiresolution approaches are more suitable for CCD/SAR image fusion than other ones proposed in this paper and the objective performance evalution of CCD/SAR image fusion approaches are effective. Foundation item: Under the auspices of Astronautical Innovation Fund of China. Biography: WANG Yan-li (1977 – ), female, a native of Xi’an City, Shaanxi Province, Ph. D. candidate of School of Automation Science and Electrical Engineering, Beijing University of Aeronautics and Astronautics, specialized in scene matching, image registration and fusion.     

基于双树复小波变换图像融合方法

基于离敞小波变换的图像融合算法在特征表达上存在不足,将双树复小波变换引入到彩色图像融合中,对彩色图像的3通道分别进行融合,各通道低频取加权平均,高频去较大值的方法.融合效果评价除了使用均方根误差、峰值信噪比和熵外,还使用了通用的主观与客观相结合的方法.实验结果表明,获得的融合图像效果较好,优于离散小波融合算法.     

遥感影像像元级融合方法与试验评价分析

不同遥感影像有着各自的优点和局限,遥感影像融合就是为了充分利用它们的优点进行互补得到新的影像。采用PCA、Brovey和小波变换三种方法对SPOT多光谱影像和全色影像进行了像元级的影像融合,小波变换采用的Mallat算法和双正交小波函数,双正交小波函数的优点是小波逆变换时不会导致相位失真,论文对融合结果进行了光谱质量和空间质量两方面的定性和定量的评价。结果表明,小波变换融合影像是三种变换中最优的,同时在小波变换中采用不同的小波基函数的融合结果在光谱质量和空间质量方面也有很大的差异,可根据实际需要选用合适的小波基函数。     

一种以CCD拼接结构的航空相机智能调焦方法

 航空相机在执行对地观测任务时, 需要精确地调焦控制系统实时调整镜头焦距以获得清晰的图像。本文以航空成像运行模式下的CCD拼接结构的特点,提出了一种智能的航空相机小波调焦方法。将图像处理的清晰度评价方法引入航空相机调焦系统,并考虑人类视觉系统(HVS)的特点设置权值,使评价效果更符合人类的主观感受。本文分析了小波基函数的性质,通过比较得出了采用symlet2小波进行三层分解构造的清晰度评价函数,对不同细节丰富程度的图像具有适应性,得到了灵敏度鲁棒性好的评价函数。并选取细节丰富的区域作为评价函数,详细地介绍了结合航空成像设备CCD拼接结构的特点,进行智能调焦的实现原理及方法。理论分析及仿真试验证实了该算法的可行性及有效性。相比传统评价方法具有较高的灵敏度,小波调焦方法的效率能够满足航空相机实时调焦的需求。     

ENVISAT ASAR数据与CBERS-02数据的融合与应用

本文首次利用ENVISAT ASAR交叉极化模式数据与中巴地球资源卫星02星(CBERS 02)CCD数据进行融合,对融合后的图像进行分析和讨论。通过对IHS变换融合、主成分分析融合和小波变换融合比较分析发现,采用IHS变换融合具有更好的目视判读效果,但是由于ASAR数据受到相干斑噪声的影响,在滤波去噪后,使融合图的空间分辨率变低。另外,该融合图包含了地物的VV极化和HH极化后向散射特征,使地物结构形态、复介电常数和地物的光谱信息等重要的地物参数信息都包含在融合图中,增强了区别地物的能力,就农作物的目视解译和分类而言,精度得到了较大的提高。     

基于小波变换的SAR图像噪声滤除方法

合成孔径雷达(SAR)的相干斑噪声严重影响图像质量,降低图像的可判读性。常用的空间自适应滤波方法在滤除噪声的同时,损失了图像中的大量边缘细节信息。介绍了一种基于小波变换的SAR图像噪声消除方法,充分考虑噪声的统计特征,并把小波变换与空间滤波两者有机结合起来。通过滤波实验与其它滤波方法的对比。表明此小波滤波方法能更有效地消除SAR图像中的斑点噪声,而且能有效地保持图像中的纹理细节和边缘信息．     

网络环境下的遥感影像金字塔纹理压缩算法与实验

 建立不同分辨率的遥感影像金字塔模型是现阶段虚拟地球平台的主要技术手段,遥感影像数据的高效压缩是模型应用的基础。论文在分析多种图像压缩技术的基础上,提出以小波变换的JPEG2000标准来压缩网络环境中遥感影像金字塔的纹理瓦片数据。首先,介绍了小波变换的JPEG2000标准的基本原理和EBCOT算法特点,然后,结合实例实现了网络环境中遥感影像金字塔纹理压缩/解压缩的具体过程,对4幅影像图像进行5级离散小波变换,分别对在1024×1024、512×512、256×256、128×128、64×64不同分块大小和1∶15、1∶30、1∶60不同压缩倍率情况下进行图像压缩和重构耗时实验,通过改变分块大小参数和压缩倍率,对压缩性能进行了对比分析,最后,从主观视觉质量和客观辐射质量,对JPEG2000压缩、DXT压缩和JPEG压缩几种方法的压缩图像进行了评价与对比分析。实验结果表明,小波变换的JPEG2000具有高质量、高压缩率、良好的抗误码能力等特性,重构图像的视觉质量与原始图像相比人眼看不出失真,而峰值信噪比较好,是一种简单有效、易于快速实现的压缩方法,更适合于网络环境下空间遥感图像数据的近无损实时压缩。     

基于MPI的大规模遥感影像金字塔并行构建方法

影像金字塔是实现影像数据多分辨率组织的重要方式,是提高影像可视化性能的有效手段。传统串行金字塔构建算法,对大规模影像数据的构建性能已无法满足遥感影像快速浏览的预处理需求。故此,其成为一个亟待解决的问题,而利用多核、多节点的高性能集群计算环境和并行机制是一个重要的技术途径。本文在共享外存的高性能集群环境下,提出使用消息传递接口(MPI)的金字塔并行构建算法,对构建遥感影像金字塔过程中的重采样与I/O 过程进行并行处理,大大缩短了遥感影像金字塔构建时间。实验结果表明:(1)该算法比传统串行构建方法的加速效果明显,对于单波段遥感影像,其加速效果可达到GDAL的5 倍以上,而对于多波段遥感影像,加速效果可达到GDAL的2 倍以上;(2)遥感影像数据量越大,并行构建算法加速效果越显著,对于大规模的遥感影像,本文提出的金字塔并行构建算法的速度可达到GDAL的10 倍左右。     

ETM与SPOT影像融合方法比较和评价——以庐山及其周边区域为例

高空间分辨率影像与多光谱影像融合,为影像处理提供了更加丰富的资源,有利于进行影像的分析和制图。本文以庐山及其周边区域的2009年5月11日Landsat ETM+与2009年5月6日SPOT 5影像数据为例,对典型的融合方法进行了对比评价。融合方法包括加权融合法,IHS变换,改进的IHS变换,高频调制融合法(频域采用Butterworth滤波器,指数滤波器),HSV变换,Brovey变换,CN变换,PC变换,Gram-Schmidt变换11种方法。评价指标为相关系数、信息熵、标准差、平均梯度、相对偏差、标准偏差。结果表明,不同的融合结果突出的侧重点不同。加权融合法、Gram-Schmidt变换,PC变换方法突出了影像的信息量丰富度、清晰度,提高空间分辨率能力。CN波谱锐化,改进的IHS变换,BW0高频调制融合,指数高频调制融合方法光谱保真度较好。改进的IHS变换,BW0高频调制融合,CN波谱锐化,指数高频调制融合法在空间分辨率及光谱保真度、清晰度、信息丰富度方面的能力较均衡。     

一种由粗到精的光学与SAR遥感图像配准算法

由于光学遥感图像和SAR图像具有明显的非线性强度差异,且SAR图像存在斑点噪声,使得其配准存在较大难度。为此,本文结合基于特征和基于区域图像配准方法的优点,并组合为混合模型,提出一种由粗到精的自动配准算法。以光学遥感图像和SAR图像分别为参考图像和待配准图像,先以基于特征点的SAR-SIFT完成粗配准,再以基于区域的ROEWA-HOG完成精配准。① 采用SAR-SIFT算法进行特征点检测和特征匹配来计算图像的仿射变换模型,以消除参考图像和待配准图像之间明显的旋转、尺度和平移差异,至此完成图像粗配准;② 在此基础上利用分块Harris角点检测在参考图像上获得特征点,并根据特征点确定待配准图像上的同名点搜索区域;③ 计算图像的ROEWA梯度,构造以特征点为中心的模板区域内的HOG特征向量,以SSD作为相似性测度搜索待配准图像上的同名点,完成高精度的图像配准;④ 进行图像配准实验,对配准结果进行目视检查和精度评估。经过多组光学与SAR图像配准实验,验证本文算法能够结合基于特征和基于区域的图像配准方法的优点,较好地抵抗光学与SAR图像之间的非线性强度、旋转、尺度、平移差异和SAR图像的噪声影响,并逐步提高配准精度,最终配准精度达到1个像素左右,实现了光学与SAR图像的高精度自动配准,能够满足光学与SAR图像后续综合应用。