遥感图像云检测方法综述

光学遥感图像中云层会对地面信息进行不同程度的遮挡,造成了地表观测信息的模糊和缺失,极大地影响遥感图像的成像质量。因此,对遥感图像中云层覆盖的检测和评估是进一步分析和利用遥感图像信息的基础和关键。通过充分的调研和对比总结,梳理了20世纪90年代以来,国内外基于遥感图像的云检测方法的发展趋势和代表性工作。将基于遥感图像的云检测方法分为三类：基于光谱阈值的方法、基于经典机器学习的方法以及基于深度学习的方法。总结了当前国内外云检测公开数据集,并对比了部分代表性工作的云检测精度。此外,简要梳理了与云检测相关的云雾(霾)检测、云雪检测、云阴影检测以及云去除等方法。对当前云检测相关工作中存在的问题和未来的发展趋势进行了分析和展望。     

基于卫星遥感的岛礁影像多类目标智能化提取

卫星遥感影像具有背景复杂、目标尺度不一、观测方向各异、纹理不清晰等特点,主流的深度学习目标检测算法不能直接适用于卫星遥感影像的目标检测。改进了RetinaNet,使其适用于卫星遥感影像。首先设计了一种新的特征融合方式,融合ResNet50输出的特征图,使得融合后的特征图同时具有高层语义信息和低层纹理细节信息。为了减弱遥感影像复杂背景对目标特征的影响,设计了特征感知模块,在减弱噪声对特征图影响的同时增强有用特征。挑选DOTA数据集中船只、飞机和存储罐图像进行训练和测试。改进的算法与RetinaNet相比,飞机、船只和存储罐的平均精度分别提高了41%、25%、24%。基于高分二号卫星（GF 2）真实影像数据的试验结果表明,提出的算法能够用于卫星遥感岛礁影像的多类目标智能化提取。     

Advances on Technology and Application of Ocean Color Remote Sensing in China (2004 - 2006)

China has great progress in the technology and application of ocean color remote sensing during 2004-2006. In this report, firstly, four major technical advances are displaying, including (1) the vector radiative transfer numerical model of coupled ocean-atmosphere system; (2) the atmospheric correction algorithm specialized on Chinese high turbid water; (3) systematical research of hyper-spectrum ocean color remote sensing; (4) local algorithms of oceanic parameters, like ocean color components, ocean primary productivity, water transparency, water quality parameters, etc. On the foundation of technical advances, ocean color remote sensing takes effect on ocean environment monitoring, with four major kinds of application systems, namely, (1) the automatic multi-satellites data receiving, processing and application system; (2) the shipboard satellite data receiving and processing system for fishery ground information; (3) Coastal water quality monitoring system, integrating satellite and airborne remote sensing technology and ship measurement; (4) the preliminary system of airborne ocean color remote sensing application system. Finally, the prospective development of Chinese ocean color remote sensing is brought forward. With Chinese second ocean color satellite (HY-1B) orbiting, great strides will take place in Chinese ocean color information accumulation and application.     

面向高分辨率遥感影像的地形辐射校正方法

地形辐射校正对获取准确的地表定量遥感精度意义重大。针对传统地形辐射校正模型不适用于高分辨率遥感影像的问题,提出了一种基于辐射传输模型,同时严控误差源的地形辐射校正方法,以资源三号01星高分辨率全色及多光谱遥感影像为例进行相关实验,实现对高分辨率遥感影像的地形辐射校正,并进行了主客观分析与评价。分析结果表明:本文提出的地形辐射校正模型和方法,能有效解决全色遥感影像在绝对辐射定标系数缺失情况下校正效果差以及如何保持高分辨率遥感影像细节等难点,较传统方法更适用于高分辨率遥感影像。      

Object-based rice mapping using time-series and phenological data

Remote sensing techniques are often used in mapping rice, but high quality time-series remote sensing data are difficult to obtain due to the cloudy weather of rice growing areas and long satellite revisit interval. As such, rice mapping is usually based on mono-temporal Landsat TM/ETM+ data, which have large uncertainties due to the spectral similarity of different vegetation types. Moreover, conventional pixel-based classification method is unable to meet the required accuracy for rice mapping. Therefore, this study proposes a new strategy for mapping rice in cloud-prone areas using fused data of Landsat-8 OLI time-series and phenological parameters, based on the object-based method. We determine the critical growth stages of paddy rice from observed phenological data and MODIS-NDVI time-series data. The spatial and temporal adaptive reflectance fusion model (STARFM) is used to blend the MODIS and Landsat data to obtain a multi-temporal Landsat-like dataset for classification. Finally, an object-oriented algorithm is used to extract rice paddies from the Landsat-like, time-series dataset. The validation experiments show that the proposed method can provide high accuracy rice mapping, with an overall accuracy of 92.38% and a kappa coefficient of 0.85.     

模态综合在遥感卫星颤动响应分析中的应用

高分辨率遥感卫星对地观测时,星上部件的运动使星体产生颤动响应,该响应会对成像质量产生一定影响。介绍了基于混合界面综合方法的遥感卫星动力学模型及其简化计算方案;给出了运动部件拢动影响下遥感卫星的颤动响应分析仿真结果。这些结果可供卫星总体部门参考。     

一种新的遥感影像清晰度评价方法

针对传统清晰度评价算法很难准确评价遥感卫星影像清晰度的问题,结合工程应用及遥感影像特点提出了一种新的遥感图像清晰度评价算法--自检测灰度梯度函数清晰度评价算法。把评价过程分为目标区域检测和清晰度特征参量提取,为解决遥感影像数据量大且景物密度与分布特点各异的问题,首先通过检测算子对一幅遥感影像中各区域进行检测来得到景物丰富且边缘明显的目标区域,然后再对目标区域进行灰度预处理并提取目标区域的边缘灰度梯度来评价清晰度。通过三组不同类型影像对该算法进行验证,分别为遥感相机在轨离焦仿真影像和噪声仿真影像,以及在轨型号遥感卫星影像,对比几种传统典型清晰度评价算法和自检测灰度梯度函数的评价效能,结果表明：文中方法满足遥感影像清晰度评价的基本要求,解决了传统算法无法横向比较不同遥感相机影像及不同大小影像清晰度的问题,是一种适合卫星遥感影像清晰度评价的方法。     

基于自适应注入模型的遥感图像融合方法

遥感图像融合的目的是融合高光谱分辨率、低空间分辨率的多光谱（MS）图像和高空间分辨率、低光谱分辨率的全色（PAN）图像,得到高光谱分辨率与高空间分辨率的融合图像。遥感图像的注入模型中如何确定注入细节及注入系数是该技术研究的关键。针对注入细节优化,先通过模拟MS传感器的特性来定义一种多尺度高斯滤波器,再用该滤波器卷积PAN图像以提取细节,得到与MS图像高度相关的细节。针对注入系数优化,综合考虑光谱信息与细节信息提出一种自适应的注入量系数。为更好地保留边缘信息,提出一种新的边缘保持权重矩阵,实现光谱信息与空间的双保真。将优化后的注入系数与注入细节相乘注入到上采样后的MS图像中,得到融合结果。对所提方法进行性能分析,并在各卫星数据集上进行大量测试,与一些先进的遥感图像融合方法进行对比,实验结果表明,所提方法在主观与综合客观指标上都能达到最优。      

星用广角线阵CCD相机的机载试验

在卫星遥感中利用广角线阵ＣＣＤ相机对地球资源信息进行“粗查“，并同时利用高分辨率ＣＣＤ相机进行“详查”的遥感器组合是一种可行的方案。文中介绍广角ＣＣＤ相机的研制及利用小型无人机进行机载试验的结果。在方案设计中主要讨论几保分辨率和辐射分辨率两项指标，也兼顾讨论诸如扫掠宽度和覆盖周期等问题。为获得好的机载试验效果将相机安装在双轴稳定平台上。     

卫星遥感图像信息作为姿态敏感器的应用研究

陆标敏感器可以从卫星遥感图像信息中提取卫星的姿态信息,它根据卫星实时图像与基准图像之间的偏移量,计算出卫星的姿态。对陆标敏感器的总体方案进行了设计,在图像匹配技术中采用基于区域特征的先粗匹配、后精匹配的匹配算法,在姿态确定算法中对可能达到的姿态测量精度进行了理论分析,研究表明陆标敏感器可以获得较高精度的卫星姿态信息,作为新型光学姿态敏感器具有重要的研究意义。     

改进的深度神经网络下遥感机场区域目标检测

卫星遥感监测器下的机场区域多类目标检测在实际生活中有着重大的军用和民用意义。为了有效提升机场区域遥感图片的检测精确率,以主流目标检测方法中更快的区域卷积神经网络（Faster R-CNN）为基础框架,针对数据侧提出了ReMD数据增强算法。同时使用更具深度的残差神经网络（ResNet）以及特征融合部件-特征金字塔网络（FPN）来提取机场区域目标更鲁棒的深层区分性特征。在末端检测网络添加新的全连接层并根据目标的类间关联性组合softmax分类器以及4个logistic regression分类器进行机场区域多类目标的精确分类。实验结果表明:相比原网络改进后的网络带来了11.6%的多类平均检测精确率的提升,达到了80.5%的mAP,与其他主流网络进行对比也有更好的精确率;同时通过适当减小建议区域的输入量,可以在降低3.2%精确率的前提下将0.512 s的检测时间提速3倍,至0.173 s,根据具体任务可以合理权衡精确率和检测速度,体现了该网络的有效性以及实用性。      

遥感图像飞机目标高效搜检深度学习优化算法

为了实现大幅面遥感图像中飞机目标的高效检测与准确定位,通过深度神经网络（DNN）的级联组合,提出了一种新颖的搜寻与检测相集成的飞机目标高效检测算法。首先,运用高性能的端到端DNN网络,进行停机坪与跑道区域的像素级高效精准分割,从而大幅度缩小飞机目标的搜索范围,以降低虚警发生概率,完成飞机目标候选检测区域的快速搜寻。然后,针对分割所得停机坪与跑道区域,借助手工数据集对YOLO网络模型进行迁移式强化训练,一方面可以弥补训练集在样本类型与数据规模上的不足,另一方面借助YOLO网络的强时效性优势对飞机目标的位置进行回归求解,可以显著提高飞机目标的检测效率。停机坪与跑道区域分割DNN网络在分割精度与时效性上具有显著优势,而迁移式强化训练YOLO网络不仅具有很高的检测效率,在检测精度上也能保持良好的性能。通过一系列综合实验与对比分析,验证了提出的搜寻与检测相集成的DNN级联组合式飞机目标高效检测算法的性能优势。      

基于策略的遥感卫星管控方法研究

针对遥感卫星管理控制中业务应用与底层调度紧耦合导致的操作复杂、人工干预多等问题,提出一种基于策略的遥感卫星管控方法。在搭建策略管控模型的基础上,采用分层控制闭环结构,构建卫星控制闭环;提出卫星集群分域管理方法,并设置策略部署机制;详细论述了基于策略的遥感卫星管控实现方法;以海上失联舰船搜救任务为典型案例进行仿真,验证该方法可以完成异构卫星集群基于事件驱动的闭环控制。基于策略的管理方法能够有效降低遥感卫星管控业务的复杂度,实现遥感卫星集群的自动控制。     

A modern robust approach to remotely estimate chlorophyll in coastal and inland zones

The chlorophyll concentration of a water body is an important proxy for representing the phytoplankton biomass. Its estimation from multi or hyper-spectral remote sensing data in natural waters is generally achieved by using (i) the waveband ratioing in two or more bands in the blue-green or (ii) by using a combination of the radiance peak position and magnitude in the red-near-infrared (NIR) spectrum. The blue-green ratio algorithms have been extensively used with satellite ocean color data to investigate chlorophyll distributions in open ocean and clear waters and the application of red-NIR algorithms is often restricted to turbid productive water bodies. These issues present the greatest obstacles to our ability to formulate a modern robust method suitable for quantitative assessments of the chlorophyll concentration in a diverse range of water types. The present study is focused to investigate the normalized water-leaving radiance spectra in the visible and NIR region and propose a robust algorithm (Generalized ABI, GABI algorithm) for chlorophyll concentration retrieval based on Algal Bloom index (ABI) which separates phytoplankton signals from other constituents in the water column. The GABI algorithm is validated using independent in-situ data from various regional to global waters and its performance is further evaluated by comparison with the blue-green waveband ratios and red-NIR algorithms. The results revealed that GABI yields significantly more accurate chlorophyll concentrations (with uncertainties less than 13.5%) and remains more stable in different waters types when compared with the blue-green waveband ratios and red-NIR algorithms. The performance of GABI is further demonstrated using HICO images from nearshore turbid productive waters and MERIS and MODIS-Aqua images from coastal and offshore waters of the Arabian Sea, Bay of Bengal and East China Sea.     

多场景高分辨率遥感图像快速拼接技术

光学遥感卫星拍摄过程中大视场与高分辨率间的相互矛盾,引起国内外研究人员的广泛关注。提出一种基于相机参数保持的快速拼接算法。该方法对高分辨率图像进行降采样处理,利用降分辨率图像的配准信息恢复相机参数,能够改进拼接过程,提高拼接效率。实验结果表明,相对于传统拼接方法,该算法能够将配准时间缩短2倍以上,将遥感图像的拼接质量提升5%~35%, 适用于建筑、森林、农田、湖泊四类场景。通过对存在噪声、亮度差异及不同模糊程度的遥感图像进行拼接实验,验证了该方法的鲁棒性和可行性。     

GNSS遥感探测卫星星座设计

随着全球导航卫星系统(Global Navigation Satellite System,GNSS)掩星大气探测技术的兴起,GNSS遥感探测数据在气象数据资源中逐步占据重要地位,但是目前的掩星探测数量远不能满足数值天气预报等应用的需求,未来更需要充分利用GNSS信号资源,开展更大规模的GNSS掩星卫星星座探测.本文以...     

Automatic lineament extraction in a heavily vegetated region using Landsat Enhanced Thematic Mapper (ETM+) imagery

Lineament extraction from satellite remotely sensed data has been one of the widely used applications of remote sensing in geology. In fact, recent advances in digital image processing allow such lineament extraction to be accomplished in semi-automatic to fully automatic approaches. However, satellite remotely sensed data acquired in heavily vegetated regions such as tropical rainforest, are vulnerable to higher inherent noise levels attributed to the resultant effects of scattering by clouds and adjacency effects of highly inhomogeneous vegetation cover within the pixel dimension. In this study, we examined the effects of noise levels to lineament extraction using a fully automatic approach, consisting of a combination of edge-line detection algorithms. Ancillary information from a digitized topographic map and image classification was used to discriminate between cultural and natural lineaments from the extracted lineaments. Adapting the combination of edge detection and a line-linking algorithm, we have found the optimal parameters for automatic lineament extraction of such complex areas using Enhanced Thematic Mapper (ETM+) data. A noise level of 30% is the maximum threshold before artifacts are generated. It is therefore concluded that the combination of edge-based and line-linking digital image processing operations with the priori local optimal parameters is crucial in lineament feature extraction in heavily vegetated regions.     

Distribution of chlorophyll and harmful algal blooms (HABs): A review on space based studies in the coastal environments of Chinese marginal seas

Monitoring of spatial and temporal distribution of chlorophyll (Chl-a) concentrations in the aquatic milieu is always challenging and often interesting. However, the recent advancements in satellite digital data play a significant role in providing outstanding results for the marine environmental investigations. The present paper is aimed to review ‘remote sensing research in Chinese seas’ within the period of 24 years from 1978 to 2002. Owing to generalized distributional pattern, the Chl-a concentrations are recognized high towards northern Chinese seas than the southern. Moreover, the coastal waters, estuaries, and upwelling zones always exhibit relatively high Chl-a concentrations compared with offshore waters. On the basis of marine Chl-a estimates obtained from satellite and other field measured environmental parameters, we have further discussed on the applications of satellite remote sensing in the fields of harmful algal blooms (HABs), primary production and physical oceanographic currents of the regional seas. Concerned with studies of HABs, satellite remote sensing proved more advantageous than any other conventional methods for large-scale applications. Probably, it may be the only source of authentic information responsible for the evaluation of new research methodologies to detect HABs. At present, studies using remote sensing methods are mostly confined to observe algal bloom occurrences, hence, it is essential to coordinate the mechanism of marine ecological and oceanographic dynamic processes of HABs using satellite remote sensing data with in situ measurements of marine environmental parameters. The satellite remote sensing on marine environment and HABs is believed to have a great improvement with popular application of technology.