Landsat TM遥感影像中厚云和阴影去除

提出了一种新的利用多时相Landsat TM影像数据进行的厚云及其阴影去除的方法。该方法通过分析厚云及其阴影的光谱特征, 设计了厚云和云阴影识别模型。该算法的实现是采用图像配准技术、非监督分类、像元替换等运算, 计算出厚云和云阴影区域的TM影像替换数据, 进而得到消除或者减少云影响的TM遥感影像。试验结果表明本文提出的厚云及其阴影去除方法效果很好, 能消除或者弱化云对TM影像数据的影响。     

多光谱遥感影像中云影区域的检测与修复

提出了一种有效针对多光谱遥感影像的云影检测与阴影区域修复方法。基于同一地区时相相近的两幅影像,充分利用碎云及阴影的光谱特性分别对云影区域进行融合增强,然后采用Otsu算法求解最佳阈值自动检测出云及阴影区域,根据云影的出现会引起两幅影像局部相应区域明显的亮度变化,可排除亮地物和水体的影响,建立归一化的云影密度图,在此基础上,采用线性加权组合与光谱直方图匹配相结合的方法对其加以修复,利用SPOT 4影像进行的实验表明其修复效果完全能够满足应用需要。     

Use of Markov Random Fields for automatic cloud/shadow detection on high resolution optical images

In this study, we propose an automatic detection algorithm for cloud/shadow on remote sensing optical images. It is based on physical properties of clouds and shadows, namely for a cloud and its associated shadow: both are connex objects of similar shape and area, and they are related by their relative locations. We show that these properties can be formalized using Markov Random Field (MRF) framework at two levels: one MRF over the pixel graph for connexity modelling, and one MRF over the graph of objects (clouds and shadows) for their relationship modelling. Then, we show that, practically, having performed an image pre-processing step (channel inter-calibration) specific to cloud detection, the local optimization of the proposed MRF models leads to a rather simple image processing algorithm involving only six parameters. Using a 39 image database, performance is shown and discussed, in particular in comparison with the Marked Point Process approach.     

资源三号测绘卫星自动云检测

光学卫星遥感影像自动云检测是卫星产品生产系统的一个重要环节。利用资源三号卫星编目生成的浏览图,采用树状判别结构进行云检测,对浏览图进行分块,提取子块图像的特征进行云地分类。由于云类和地物类过于繁杂,且浏览图的分辨率较低,传统通过图像特征对云地进行分类的算法有很大的局限性,针对这一问题,本文提出了在分类之前对原始的子块图像进行增强处理,扩大云和地物的纹理差异,然后以二阶矩、一阶差分等作为云地分类的图像特征,并在多尺度空间内进行特征延拓,经过综合分析估计云在影像中的比例。该云检测算法应用于资源三号卫星应用系统工程,实际测试结果表明,该算法能够较好地提升云量检测的准确率。     

基于支持向量机的遥感影像厚云及云阴影去除

本文提出了一种基于支持向量机的遥感影像厚云及云阴影去除方法。首先利用支持向量机的学习性能检测影像中的云层,并利用太阳角度信息,判定云阴影区域,得到云层和云阴影的二值图。再对影像进行支持向量值轮廓波变换,利用云层和云阴影二值图生成的选择矩阵,对变换系数进行多层镶嵌,完成云层及云阴影的初去除。对影像镶嵌未能去除的云层及云阴影,通过统计学补偿的方法进行修复。最后重构图像并进行中值滤波实现厚云及云阴影去除。仿真实验表明,该方法能更好地再现云层覆盖区域的地物信息,去云后的图像具有更好的光滑度和清晰度。     

变序字典学习AO-DL的资源三号遥感影像云去除

采用了一种压缩感知方法进行遥感影像去云。该方法以压缩感知为理论基础,在采用K-SVD字典学习与稀疏表示的正交匹配追踪算法(OMP)相结合的同时,在字典原子训练的过程中加入某种特定的排序规则,使得各个影像字典在拥有各自影像属性的同时其原子也具备相似的排列顺序,减小影像间差异的干扰,使得遥感影像受云和阴影污染区域的重建取得良好的效果。最后应用两组相同地区不同时域的资源三号卫星影像进行了试验验证。     

基于显著性图像与纹理特征的遥感影像云检测

遥感影像云检测是遥感影像处理中非常关键的环节,准确识别影像含云区域能够提升影像的利用价值。根据遥感影像的成像特点,将阈值法和纹理特征结合实现云和下垫面的分割。首先将影像从RGB(red-green-blue)空间转化为HSI(hue-saturation-intensity)空间,进而构建影像的显著性图像,利用Otsu法对显著性图像进行粗分割,再基于灰度共生矩阵分析云和下垫面的纹理特征,进一步提取出准确的云区。实验表明,该算法复杂度较低,提取效果良好。     

融合多尺度视网膜图像增强的动态云检测

云检测是气象卫星各类定量遥感产品的基础,无论是以云图为基础的天气分析还是以去云为前提的各类大气和地表参数反演、沙尘火情等灾害检测,都需要对遥感影像中的云进行准确识别,尤其是薄云和云边缘等细节识别。针对静止气象卫星(以Himawari-8为例)精细化云检测,本文提出了一种基于多尺度视网膜图像增强的动态云检测算法。该算法基于云层与背景信息辐射特征不同的原理,构建可见光和红外波段的晴空辐射背景场,通过多尺度图像增强和最大类间差方法对辐射差值进行云细节信息的增强和提取。利用2021-2022年的75景MODIS云检测产品作为验证数据进行算法精度验证,整体上算法精度达到91.13%,召回率为94.02%,精确率为86.71%,有较强的适用性和稳健性,且已经较好地支撑了近两年的定量遥感产品业务化应用。     

基于多维云空间的多光谱遥感影像边缘检测研究

在云模型的基础之上,提出一种基于多维云空间的遥感影像边缘检测算法。该方法依据矢量角相似性准则并结合邻域关系进行图像区域生长,在此基础之上根据影像不同波段的数据特征建立多维云空间映射模型,将数据从图像空间映射到云空间。通过多维云的逻辑运算生成边界云并进行向量综合。利用云的数字特征反推隶属度,构建边缘模糊特征平面,并通过最大模糊熵原则确定最优阈值,对图像模糊边界进行提取。试验结果表明,该算法在多光谱遥感影像中能取得较好检测效果。     

结合超像素和卷积神经网络的国产高分辨率遥感影像云检测方法

由于国产高分辨率卫星遥感影像波段少、光谱范围窄，导致传统云检测方法精度低。本文提出了基于卷积神经网络的高分辨率遥感影像云检测方法。首先采用主成分分析非监督预训练网络结构，获取待测遥感影像云特征；然后采用超像素分割方法进行影像分割；最后将检测结果影像块拼接，完成整幅影像云检测。试验效果评价表明，基于卷积神经网络的高分辨率遥感影像云检测方法不受光谱范围限制，云检测精度高，误判较少，适合国产高分辨遥感影像云检测。     

融合引导滤波和迁移学习的薄云图像中地物信息恢复算法

薄云覆盖遥感图像使图像上的地物信息模糊。本文给出了一种融合引导滤波和迁移学习的薄云图像中地物信息恢复算法。首先利用多方向非抽样对偶树复小波变换对薄云目标图像和无云引导图像进行多分辨率分解,再对分解后的低频子带分别进行支持向量引导滤波和迁移学习,对分解后的高频子带利用修正的Laine增强函数进行增强,然后应用基于区域能量的选择和加权相结合的方法对引导滤波输出和迁移学习模型预测的低频子带进行融合,最后对增强后的高频子带和融合后的低频子带进行多方向非抽样对偶树复小波逆变换重构,获得地物信息恢复图像。Landsat-8 OLI多光谱图像的试验结果表明,支持向量引导滤波能够有效保留目标图像的地物细节信息,域自适应的迁移学习能有效扩展可利用的多源多时相遥感图像范围,通过融合引导滤波和迁移学习能有效去除遥感图像上的薄云,获得较好的地物信息恢复效果。     

增强型多时相云检测

针对云检测在高亮度地表以及雪覆盖区域存在过度检测的问题,设计了一种不依赖热红外波段的增强型多时相云检测EMTCD(Enhanced Multiple Temporal Cloud Detection)算法。首先,利用云的光谱特征建立单时相云检测规则,并基于云、雪的光谱差异构建了增强型云指数ECI(Enhanced Cloud Index),改进了云、雪的区分能力;其次,以同一区域无云影像为参考,基于ECI指数构建了多时相云检测算法,较好地克服了单时相云检测中高亮度地表、雪和云容易混淆的问题,提高了云检测的精度;最后,选择两个典型区域的Landsat-8 OLI影像,对比分析了不同算法的云检测结果。实验结果表明:ECI指数能够有效区分云、雪,EMTCD方法的平均检测精度达到93.2%,高于Fmask(Function of mask)(81.85%)、MTCD(Multi-Temporal Cloud Detection)(66.14%)和Landsat-8地表反射率产品LaSRC(Landsat-8 Surface Reflectance Code)的云检测结果(86.3%)。因此,本文提出的EMTCD云检测算法能够有效地减少高亮度地表和雪的干扰,实现不依赖热红外波段的高精度云检测。     

辐射特征支持下的城市高分影像阴影校正

大量城市建筑使得高分影像中含有许多阴影区。这些阴影区在土地利用分类、植被绿度调查等遥感应用中会较大地影响结果精度,降低数据使用效率并增加研究成本。基于同一地物阴影区与临近非阴影区反射率相等这一辐射特征关系,通过建立辐射传输方程,发展了一种新的城市高分遥感影像阴影校正方法 RERB(Reflectance Equality Relationship Based Method)。利用RERB对不同城市(北京和荷兰Enschede)不同高分多光谱影像(Geo Eye-1和Quick Bird)进行阴影校正,并对比分析其与被广泛采用的均值方差变换法MVT(Mean and Variance Transformation)的校正结果,通过定性和定量精度评价发现:(1)RERB能很好地将城市阴影区影像视觉特征(颜色、纹理、色调等)信息恢复到与非阴影区同一水平上;(2)RERB恢复后的阴影区具有丰富的细节信息且在视觉上与临近非阴影区具有良好的一致性;(3)RERB恢复后的城市柏油路面和水泥路面阴影区辐射信息具有较低的误差,可见光-近红外波段的平均误差分别为7%和9%。同时RERB能较好地恢复城市阴影区植被波谱特征信息。     

高山冰川多时相多角度遥感信息提取方法

提出一种多角度遥感影像的冰川信息提取方法。通过"全域—局部"的阈值分割方法获取短时期内不同时相的遥感影像的冰雪边界,结合地形信息和多时相遥感影像的太阳角度信息,联合消除山体阴影对冰川的遮挡,并以多期影像的最小冰雪边界作为最佳冰川边界。以托木尔峰西侧冰川为研究对象,采用2009—2010年4个时相的遥感影像提取冰川信息。结果表明多角度遥感提取的冰川边界效果好,能有效地排除积雪与山体阴影的干扰。     

联合云量自动评估和加权支持向量机的Landsat图像云检测

针对ACCA（云量自动评估）算法难以检测Landsat图像中的半透明云问题,提出了一种ACCA和WSVM（加权支持向量机）相结合的云检测算法．首先根据云在不同波段中的大气辐射特点,结合Landsat ETM+图像数据的光谱特性,利用ACCA算法将图像像元初步分成云像元、非云像元和待定像元,再以云的光谱特性构造特征向量,利用WSVM算法进行待定像元的云层检测,最终获得全部图像的云检测结果．仿真实验结果表明,该方法既具有ACCA算法的云检测优势,还对ACCA算法难以识别的半透明云有很好的检测效果．     

基于多光谱分析的HJ-1B数据云检测算法

云检测是遥感数据预处理中的一个重要环节。本文通过分析云及典型地物光谱特性,结合HJ-1B传感器波段设置,提出了一种基于多光谱分析的云检测算法。该方法针对不同下垫面类型,选取合适的波段组合逐一建立云掩模。利用该算法对多幅影像进行去云处理,结果表明,该方法能够很好地检测出不同时期不同背景上空的云像元。     

An effective thin cloud removal procedure for visible remote sensing images

Clouds are obstructions for land-surface observation, which result in the regional information being blurred or even lost. Thin clouds are transparent, and images of regions covered by thin clouds contain information about both the atmosphere and the ground. Therefore, thin cloud removal is a challenging task as the ground information is easily affected when the thin cloud removal is performed. An efficient and effective thin cloud removal method is proposed for visible remote sensing images in this paper, with the aim being to remove the thin clouds and also restore the ground information. Since thin cloud is considered as low-frequency information, the proposed method is based on the classic homomorphic filter and is executed in the frequency domain. The optimal cut-off frequency for each channel is determined semi-automatically. In order to preserve the clear pixels and ensure the high fidelity of the result, cloudy pixels are detected and handled separately. As a particular kind of low-frequency information, cloud-free water surfaces are specially treated and corrected. Since only cloudy pixels are involved in the calculation, the method is highly efficient and is suited for large remote sensing scenes. Scenes including different land-cover types were selected to validate the proposed method, and a comparison analysis with other methods was also performed. The experimental results confirm that the proposed method is effective in correcting thin cloud contaminated images while preserving the true spectral information.     

加入多时相纹理的遥感变化检测

基于遥感技术的变化检测是遥感应用的一个重要方面.传统基于遥感的变化检测方法一般是利用光谱信息,较少注意多时相图像间的光谱特征相关关系分析.本文运用地统计学的伪交叉变差函数(Pseudo Cross Variogram)计算多时相图像纹理,定量表达多时相图像间的空间相关关系,并将得到的多时相纹理信息与光谱信息一起用于多时相变化检测.实验结果表明,加入多时相纹理信息可以显著提高变化检测精度,是一种有效的方法.     

Contextual Spatiospectral Postreconstruction of Cloud-Contaminated Images

A general method has been proposed recently for the contextual reconstruction of cloud-contaminated areas in multitemporal multispectral images. It is based on the idea of making the prediction process learn from information available in the cloud-free neighborhood of contaminated areas. Though promising, this method does not fully exploit all available information, thus leaving room for further methodological enhancements. This letter presents a postreconstruction methodology for improving the contextual reconstruction process by opportunely capturing spatial and spectral correlations characterizing the considered image. In addition, we propose a solution to a problem that has not yet been addressed in the remote sensing literature, i.e., the generation of an error map beside the reconstructed images to provide end-users with helpful indications about reconstruction reliability. Thorough experiments conducted on a multitemporal sequence of Landsat-7 ETM+ images are reported and discussed.     

Normalizing shadows in multi-temporal aerial frame imagery using relative radiometric adjustments to support near-real-time change detection

This study addresses the problem of shadows in multi-temporal imagery, which is a key issue with change detection approaches based on image comparison. We apply image-to-image radiometric normalizations including histogram matching (HM), mean-variance (MV) equalization, linear regression based on pseudo-invariant features (PIF-LR), and radiometric control sets (RCS) representing high- and low-reflectance extrema, for the novel purpose of normalizing brightness of transient shadows in high spatial resolution, bi-temporal, aerial frame image sets. Efficient shadow normalization is integral to remote sensing procedures that support disaster response efforts in a near-real-time fashion, including repeat station image (RSI) capture, wireless data transfer, shadow detection (as precursor to shadow normalization), and change detection based on image differencing and visual interpretation. We apply the normalization techniques to imagery of suburban scenes containing shadowed materials of varied spectral reflectance characteristics, whereby intensity (average of red, green, and blue spectral band values) under fully illuminated conditions is known from counterpart reference images (time-1 versus time-2). We evaluate the normalization results using stratified random pixel samples within transient shadows, considering central tendency and variance of differences in intensity relative to the unnormalized images. Overall, MV equalization yielded superior results in our tests, reducing the radiometric effects of shadowing by more than 85 percent. The HM and PIF-LR approaches showed slightly lower performance than MV, while the RCS approach proved unreliable among scenes and among stratified intensity levels. We qualitatively evaluate a shadow normalization based on MV equalization, describing its utility and limitations when applied in change detection. Application of image-to-image radiometric normalization for brightening shadowed areas in multi-temporal imagery in this study proved efficient and effective to support change detection.