基于FLAASH模型的MODIS1B数据的大气校正

主要介绍了FLAASH大气校正模型的主要原理及算法,并运用ENVI软件中的FLAASH大气校正模型对上海长江口地区MODIS1B卫星影像进行大气校正,对校正前后的影像进行对比分析。研究表明,MODIS1B卫星影像经过FLAASH大气校正后,较好地消除了大气影响。     

山地地区高分卫星影像正射纠正

随着国内外对地观测卫星技术的不断进步,山地地区高分卫星影像获取能力大幅提升。针对山地区域高分卫星影像云雾多、成像阴影多、纠正位置精度控制难等问题,该文提出山地区域高分卫星正射纠正生产方案。试验证明,经过大气校正、可视化的地形修复等处理后可解决薄雾去除和影像扭曲变形等问题;同时,GPU并行计算可实现影像快速融合、控制点自动选择等操作,提高生产效率,可为山地区域高分卫星影像快速处理提供参照。     

基于FLAASH与QUAC模型的SPOT 5影像大气校正比较

卫星遥感影像的大气校正是定量遥感研究的前提与难点之一,大气校正有多种方法和模型。采用FLAASH与QUAC模型对覆盖长株潭地区的SPOT 5遥感影像进行大气校正,进而对校正前后的影像进行视觉、地物光谱曲线对比分析。结果表明,两种模型有其特定的适用范围,均能基本消除大气的影响,能较好地恢复各类地物光谱的典型特征;采用FLAASH模型的精度较QUAC模型的精度高;应用QUAC模型较FLAASH简便,它对输入参数和仪器标定精度的依赖性小。     

基于神经网络的二类水体大气校正

以MERIS高光谱影像为数据源,根据现有大气传输模型和大气校正方法,探索了适合于内陆湖泊二类水体的高光谱遥感影像大气校正方法。在6S辐射传输模型的基础上,构建了基于神经网络的二类水体大气校正算法。通过构建输入卫星辐亮度直接提取离水反射率的模型,无需同步气溶胶参数,即可实现大气校正。对2010年8月9日的MERIS影像进行大气校正,并将校正后的遥感反射率与准同步实测离水反射率进行对比分析,结果表明,大气校正过程有效去除了大气效应的影响,经过大气校正的13个波段的平均相对误差分布在10%～40%,得到了与实测值相近的水体遥感反射率。     

CBERS-02卫星数据大气校正的快速算法

研究中巴资源卫星CBERS-02卫星数据大气校正的快速算法。首先假设地表为朗伯体，利用MODTRAN4．0软件模拟大气状况，讨论了不同地表反射率对大气校正参数的影响；然后计算出大气校正参数，生成查找表，根据CBERS卫星图像的不同情况确定不同的大气校正参数，从而对整幅图像进行大气校正。最后，利用核驱动模型对地表进行BRDF校正。     

高分辨率遥感影像几何纠正方法

在对某些地区的高分辨率卫星影像进行几何校正时,由于地形等因素影响难以寻找控制点,从而导致控制点较少且分布不均匀,影响了校正精度。针对这些问题,对线阵推扫式的高分辨率遥感影像的特点进行了分析,提出了利用直线这种更高级的几何特征对影像进行几何校正的方法;构建空间直线矢量,提出了基于空间直线矢量的多项式几何校正模型;并利用SPOT5高分辨率卫星影像数据对该校正模型进行精度验证。     

基于6S模型的山区环境星影像大气校正研究

针对山区地形复杂,高程变化较大的特点,利用6S辐射传输模型建立查找表,结合研究区DEM对影像进行了大气校正。试验结果表明,校正后的影像清晰度、对比度提高了,消除了大气对NDVI的负面影响。     

遥感影像地形校正物理模型的简化与改进

针对考虑太阳直接辐射、天空散射辐射以及来自于附近地表反射辐射的复杂地形EMT 遥感影像地形校正物理模型公式复杂、计算繁琐、不易实施的特点以及存在过度校正的缺陷,对该物理模型进行简化和改进。提出一套简化模型相关参数,即r值,Vt,Vd,T↓(λ,θ)等的计算方案,从而简化计算过程,提高计算效率。并针对该地形校正物理模型朗伯体假设的缺陷,引入Minnaert参数k时模型进行非朗伯体修正。简化和改进的地形校正物理模型的校正实验结果表明该模型很好地消除了复杂地形EMT 遥感影像的地形阴影,从而证明该地形校正物理模型的简化和改进方案可行。     

基于Radarsat-2数据的星载SAR正射校正和地形辐射校正方法研究

地形对雷达影像的几何和辐射特性都有强烈的影响.对雷达影像进行定量分析和参数提取之前,必须对SAR影像进行精确的几何校正和辐射校正,消除地形的影响.基于RD定位模型和数字高程模型建立一种正射校正和地形辐射校正(TRC)方法.通过试验,从定性和定量两方面评价正射校正和地形辐射校正结果的有效性.比较基于投影角和基于局部入射角...     

陡峭山区影像的半经验地形校正

本文提出一种新的半经验地形校正模型SCEDIL(Simple topographic Correction using Estimation of Diffuse Light),该模型通过结合DEM与光学影像数据寻找局部区域内完全光照和阴影的水平像元,并以光照、阴影水平像元的平均反射率值估算局部区域散射辐射比,提高了陡峭山区影像的地形校正精度。以高分一号卫星和Landsat ETM+影像为例,从目视判读和定量分析两个方面,比较分析该算法与传统半经验地形校正算法(C、SCS+C)的校正结果。结果表明:(1)对较为平坦的地形,SCEDIL和C、SCS+C校正都有较好的目视结果;对地面起伏较大的陡峭地形,C、SCS+C校正后,原阴影区域易呈现破碎化特征,SCEDIL校正后,原阴影区域过渡较为平滑。(2)SCEDIL校正后,各波段反射率的均值和标准差优于C、SCS+C校正,SCEDIL校正后,影像总分类精度与同类地物光谱信息均一性均优于C和SCS+C校正。SCEDIL半经验地形校正方法能有效地去除影像中的地形干扰,尤其对陡峭地形的校正效果,优于常规地形校正模型。     

Use of Field Spectroscopy for Exploring the Impact of Atmospheric Effects on Landsat 5 TM/7 ETM+ Satellite Images Intended for Hydrological Purposes in Cyprus

The spectroradiometric retrieved reflectance of a local crop, namely, beans (Phaseolus vulgaris), is directly compared to the reflectance of Landsat 5TM and 7ETM+ atmospherically corrected and uncorrected satellite images. Also, vegetation indices from the same satellite images—atmospherically corrected and uncorrected—are compared with the corresponding vegetation indices produced from field measurements using a spectroradiometer. Vegetation Indices are vital in the estimation of crop evapotransiration under standard conditions (ET_c) because they are used in stochastic or empirical models for describing crop canopy parameters such as the Leaf Area Index (LAI) or crop height. ET_c is finally determined using the FAO Penman-Monteith method adapted to satellite data, and is used to examine the impact of atmospheric effects. Regarding the reflectance comparison, the main problem was observed in Band 4 of Landsat 5TM and 7ETM+, where the difference, for uncorrected images, was more than 20% and statistically significant. Results regarding ET_c show that omission or ineffective atmospheric corrections in Landsat 5TM,/7ETM+ satellite images always results in a water deficit when estimating crop water demand. Diminished estimated crop water requirements can result in a reduction in output or, if critical, crop failure. The paper seeks to illustrate the importance of removing atmospheric effects from satellite images designated for hydrological purposes.     

An approach to the radiometric aerotriangulation of photogrammetric images

Harnessing the radiometric information provided by photogrammetric flights could be useful in increasing the thematic applications of aerial images. The aim of this paper is to improve relative and absolute homogenization in aerial images by applying atmospheric correction and treatment of bidirectional effects. We propose combining remote sensing methodologies based on radiative transfer models and photogrammetry models, taking into account the three-dimensional geometry of the images (external orientation and Digital Elevation Model). The photogrammetric flight was done with a Z/I Digital Mapping Camera (DMC) with a Ground Sample Distance (GSD) of 45 cm. Spectral field data were acquired by defining radiometric control points in order to apply atmospheric correction models, obtaining calibration parameters from the camera and surface reflectance images. Kernel-driven models were applied to correct the anisotropy caused by the bidirectional reflectance distribution function (BRDF) of surfaces viewed under large observation angles with constant illumination, using the overlapping area between images and the establishment of radiometric tie points. Two case studies were used: 8-bit images with applied Lookup Tables (LUTs) resulting from the conventional photogrammetric workflow for BRDF studies and original 12-bit images (Low Resolution Color, LRC) for the correction of atmospheric and bidirectional effects. The proposed methodology shows promising results in the different phases of the process. The geometric kernel that shows the best performance is the Lidense kernel. The homogenization factor in 8-bit images ranged from 6% to 25% relative to the range of digital numbers (0–255), and from 18% to 35% relative to levels of reflectance (0–100) in the 12-bit images, representing a relative improvement of approximately 1–30%, depending on the band analyzed.     

大气校正对SPOT卫星遥测水质的影响

藉由卫星遥测进行河川水质监测，目前尚没有较明确可行之方法，如何利用较为简单且适当的SPOT卫星遥测大气校正方法，正确辨识水体水质，是本研究的主要目的。利用SPOT卫星作两阶段非监督式及监督式自动分类，确认卫星影像中水质测站对应之水体样本，并将所有样本依季节分群，俾让卫星监测水体水质样本较为均质。模拟方式采用多变量回归、类神经网络及判别分析3种模式，并比较4种不同之大气校正程序。结果发现。以水质及其指标整体预测来看，类神经网络预测结果较优于多变量回归及判别分析的结果，大气校正方法以直接采用灰度值并消除最暗像元灰度值之校正方法，即可达到不错之预测结果。综合而言，以SPOT或分辨率更高之卫星光谱遥测水质是简单可行，但仍需更多数据以验证其精确度。     

北京市平原区2015年—2019年0.8 m地表反射率数据集

地表反射率产品作为最重要的定量遥感产品,是很多参量化遥感产品的基础数据源,可以被广泛应用于林业、农业、水资源、生态环境、城市环境等典型应用领域。对于米级高分辨率的遥感影像,国内外不提供反射率影像产品。目前主流的国产高分数据源大部分都是蓝、绿、红、近红4个波段的多光谱数据,缺少短波红外波段,难以满足陆地区域浓密植被算法或者清洁水体的短波红外信号近似看作零的假设条件,要完成精确大气校正、生产大范围的GF-2影像拼接的地表反射率产品,是一个挑战。为了更好的推广高分地表反射率的应用,本研究针对北京市平原区,运用全色和多光谱融合、几何精校正和相对辐射一致化算法,以Sentinel-2为参考影像,生成了一套几何精校正后的0.8 m地表反射率影像集。该数据集时间范围为2015年—2019年,每年一期,包含年度产品的覆盖情况及分布矢量,共计184景地表反射率影像,数据量总共为1.63 TB。     

Calibration of ASTER and ETM + imagery using empirical line method—A case study of north-east of Hajjah, Yemen

This study is aimed at using the Empirical Line Method (ELM) to eliminate atmospheric effects with respect to visible and near infrared bands of advanced spaceborne thermal emission and reflection radiometer (ASTER) and enhanced thematic mapper plus (ETM+) data. Two targets (Amran limestone as light target and quartz-biotite-sericite-graphite schists as dark target), which were widely exposed and easy to identify in the imagery were selected. The accuracy of the atmospheric correction method was evaluated from three targets (vegetation cover, Amran limestone and Akbra shale) of the surface reflectance. Analytical spectral devices (ASD) FieldSpec3 was used to measure the spectra of target samples. ETM+ data were less influenced by the atmospheric effect when compared to ASTER data. Normalized differences vegetation indices (NDVI) displayed good results with reflectance data when compared with digital number (DN) data because it is highly sensitive to ground truth reflectance (GTR). Most of the differences observed before and after calibration of satellite images (ASTER and ETM+) were absorbed in the SWIR region.      

航天高光谱遥感器CHRIS的水体图像大气校正

CHRIS(Compact High Resolution Imaging Spectrometer)是欧空局于2001年10月发射的PROBA-1卫星上搭载的探索性高光谱遥感器,它具备高空间分辨率、多角度观测、高光谱成像等特点,为水质遥感监测提供不可多得的数据源。基于卫星遥感图像定量监测水质,一个关键步骤就是进行精确的大气校正,提取水面反射率。相比陆地遥感图像,水面反射率是弱信号,对大气校正的要求更高。6S(Second Simulation of SatelliteSignal in the Solar Spectrum)和MODTRAN(MOderate resolution TRANsmittance code)是两种常用的大气辐射传输模型。本文选取基于6S的REMS(Remote Sensing Environmental Monitoring System)和基于MODT-RAN的ACORN(Atmospheric CORrection Now)两种大气校正软件,对太湖梅梁湾的三景不同成像角度CHRIS图像进行大气校正,将大气校正后的图像水体反射率与地面同步实测水体反射率进行比较分析。结果表明,经过大气校正的CHRIS图像得到的水面反射率与实测反射率波形十分地接近,在全部波长范围内的相关系数达到90%。分析实测的水体反射率角度特征,发现图像的角度特征更明显。三个观测角度下反射率之间的各差值都呈现出在绿光波段较大,在红光和近红外波段偏小的特点,这和实测结果相符。ACORN校正后的图像的角度特征更好地与实测结果吻合。     

Analysis of spectral characteristics among different sensors by use of simulated RS images

This research, by use of RS image-simulating method, simulated apparent reflectance images at sensor level and ground-reflectance images of SPOT-HRV, CBERS-CCD Landsat-TM and NOAA14-AVHRR's corresponding bands. These images were used to analyze sensor's differences caused by spectral sensitivity and atmospheric impacts. The differences were analyzed on Normalized Difference Vegetation Index(NDVI). The results showed that the differences of sensors' spectral characteristics cause changes of their NDVI and reflectance. When multiple sensors' data are applied to digital analysis, the error should be taken into account. Atmospheric effect makes NDVI smaller, and atmospheric correction has the tendency of increasing NDVI values. The reflectance and their NDVIs of different sensors can be used to analyze the differences among sensor's features. The spectral analysis method based on RS simulated images can provide a new way to design the spectral characteristics of new sensors.     

鄱阳湖HJ-1A/B卫星CCD影像大气校正研究

以准同步的Terra/MODIS反演的气溶胶为辅助,采用FLAASH模型对2009-10-24鄱阳湖HJ-1A/B卫星CCD影像进行大气校正处理。结果表明,大气影响可以被有效去除,在水体遥感反射率较高的红、绿波段,大气校正精度较高,平均相对误差分别为13.4%和9.8%;而在水体遥感反射率较低的近红外、蓝波段,大气校正精度较低,这可能与波段不同的信噪比和陆地邻近像元效应有关。     

Analysis of spectral characteristics among different sensors by use of simulated RS images

1　ＩｎｔｒｏｄｕｃｔｉｏｎＩｎｍａｎｙｒｅｍｏｔｅｓｅｎｓｉｎｇａｐｐｌｉｃａｔｉｏｎｓ,ｔｈｅａｎａｌｙｓｉｓｏｆｔｈｅｍｕｌｔｉ＿ｓｅｎｓｏｒｓｄａｔａｆｒｏｍｄｉｆｆｅｒｅｎｔｔｉｍｅａｎｄｄｉｆｆｅｒｅｎｔｓｅｎｓｏｒｓｉｓｏｆｔｅｎｕｓｅｄｔｏｉｍｐｒｏｖｅｔｈｅｐｒｅｃｉ ｓｉｏｎｏｆｄａｔａｐｒｏｃｅｓｓｉｎｇ .Ｅｓｐｅｃｉａｌｌｙ ,ｉｎｔｈｅｄｙｎａｍｉｃｃｈａｎｇｅ＿ｍｏｎｉｔｏｒｉｎｇｏｆｅｎｖｉｒｏｎｍｅｎｔａｎｄａｇｒｉｃｕｌｔｕｒｅｒｅｍｏｔｅｓｅｎｓｉｎｇ ,ｐｅｏｐｌ…     

Surface Reflectance Climate Data Records (CDRs) is a Reliable Landsat ETM+ Source to Study Chlorophyll Content in Pecan Orchards

We evaluated the relationships among three Landsat Enhanced Thematic Mapper (ETM+) datasets, top-of-atmosphere (TOA) reflectance, surface reflectance climate data records (surface reflectance-_CDR) and atmospherically corrected images using Fast Line-of-Sight atmospheric analysis of Spectral Hypercubes model (surface reflectance-_FLAASH) and their linkto pecan foliar chlorophyll content(chl-cont). Foliar chlorophyll content as determined with a SPAD meter, and remotely-sensed data were collected from two mature pecan orchards (one grown in a sandy loam and the other in clay loam soil) during the experimental period. Enhanced vegetation index derived from remotely sensed data was correlated to chl-cont. At both orchards, TOA reflectance was significantly lower than surface reflectance within the 550–2400 nm wavelength range. Reflectance from atmospherically corrected images (surface reflectance-_CDR and surface reflectance-_FLAASH) was similar in the shortwave infrared (SWIR: 1550–1750 and 2080–2350 nm) and statistically different in the visible (350–700 nm). Enhanced vegetation index derived from surface reflectance-_CDR and surface reflectance-_FLAASH had higher correlation with chl-cont than TOA. Accordingly, surface reflectance is an essential prerequisite for using Landsat ETM+  data and TOA reflectance could lead to miss-/or underestimate chl-cont in pecan orchards. Interestingly, the correlation comparisons (Williams t test) between surface reflectance-_CDR and chl-cont was statistically similar to the correlation between chl-cont and commercial atmospheric correction model. Overall, surface reflectance-_CDR, which is freely available from the earth explorer portal, is a reliable atmospherically corrected Landsat ETM+ image source to study foliar chlorophyll content in pecan orchards.