Revisiting hyperspectral remote sensing: origin, processing, applications and way forward
01 Jan 2020-pp 3-21
TL;DR: This chapter discusses the origin of hyperspectral remote sensing, its importance, preprocessing, inversion models suitable for hyperspectrals, as well as several possible applications, including but not limited to, vegetation analysis, agriculture, urban, water quality, and mineral identification.
Abstract: After several years of research and development in hyperspectral imaging systems that enriched our knowledge and enhanced our capacity to explore the Earth, these systems have been widely accepted by the remote sensing community. They have evolved as major techniques and have now entered the mainstream of the earth observation data users. This chapter discusses the origin of hyperspectral remote sensing, its importance, preprocessing, inversion models suitable for hyperspectral datasets, as well as several possible applications, including but not limited to, vegetation analysis, agriculture, urban, water quality, and mineral identification. The chapter concludes by looking at the way forward for hyperspectral remote sensing.
Citations
More filters
TL;DR: In this paper, two nonparametric machine learning algorithms viz Support Vector Machines (SVMs) with different kernel functions were employed for the prediction of above ground biomass using different combinations of VV, VH, Normalized Difference Vegetation Index (NDVI) and Incidence Angle (IA).
18 citations
TL;DR: This study is one of the few focusing on exploring and demonstrating the added value of the synergistic use of the Cellular Automata Markov Chain Model Coupled with Fragmentation Statistics in forest degradation analysis and prediction.
Abstract: Forest degradation is considered to be one of the major threats to forests over the globe, which has considerably increased in recent decades. Forests are gradually getting fragmented and facing biodiversity losses because of climate change and anthropogenic activities. Future prediction of forest degradation spatiotemporal dynamics and fragmentation is imperative for generating a framework that can aid in prioritizing forest conservation and sustainable management practices. In this study, a random forest algorithm was developed and applied to a series of Landsat images of 1998, 2008, and 2018, to delineate spatiotemporal forest cover status in the sanctuary, along with the predictive model viz. the Cellular Automata Markov Chain for simulating a 2028 forest cover scenario in Shoolpaneshwar Wildlife Sanctuary (SWS), Gujarat, India. The model’s predicting ability was assessed using a series of accuracy indices. Moreover, spatial pattern analysis—with the use of FRAGSTATS 4.2 software—was applied to the generated and predicted forest cover classes, to determine forest fragmentation in SWS. Change detection analysis showed an overall decrease in dense forest and a subsequent increase in the open and degraded forests. Several fragmentation metrics were quantified at patch, class, and landscape level, which showed trends reflecting a decrease in fragmentation in forest areas of SWS for the period 1998 to 2028. The improvement in SWS can be attributed to the enhanced forest management activities led by the government, for the protection and conservation of the sanctuary. To our knowledge, the present study is one of the few focusing on exploring and demonstrating the added value of the synergistic use of the Cellular Automata Markov Chain Model Coupled with Fragmentation Statistics in forest degradation analysis and prediction.
16 citations
TL;DR: The current work explored the solution to handle class imbalance by resampling the datasets before the application of classification algorithms by proposing a new computationally efficient class wise resampled technique which is based on SMOTE and centroid-based clustering.
Abstract: In the era of climate change, monitoring and effective retrieval of soil, water bodies, vegetation parameters etc. are of utmost importance which is successfully being executed using remote sensing from last few decades. The advancement of technologies has enabled us to reach effective decision making through these sensors. The advantage of acquiring multitemporal spatially continuous data sometimes turns into a disadvantage due to class imbalance where minority class instances are often misclassified by most of the classifiers. The current work explored the solution to handle this problem by resampling the datasets before the application of classification algorithms by proposing a new computationally efficient class wise resampling technique which is based on SMOTE and centroid-based clustering. The experiment was conducted on two benchmarked publicly available hyperspectral datasets. The output of the current work shows the superiority of the current work over past studies based on the performance evaluation metrics, accuracy, precision, recall and kappa values.
10 citations
TL;DR: Interband information overlapping enhances redundancy in hyperspectral data, which makes identification of application-specific optimal bands essential for obtaining accurate information about folia...
Abstract: Interband information overlapping enhances redundancy in hyperspectral data. This makes identification of application-specific optimal bands essential for obtaining accurate information about folia...
9 citations
TL;DR: In this article, the authors provided modified hyperspectral indices through detection of optimum bands for estimating species diversity within Shoolpaneshwar Wildlife Sanctuary (SWS) in India.
Abstract: Species diversity quantification is a crucial step towards the biodiversity conservation and ecosystem health. The technological advancements and existing limitations of multispectral remote sensing has increased the popularity of hyperspectral remote sensing which found its use in the estimation of species diversity. The contiguous narrow bands available in hyperspectral data enables the improvised assessment of diversity index but the overlapping of the information could result in the redundancy that needs to be handled. Due to this, the idenfication of optimal bands is very important; hence, the current study provides modified hyperspectral indices through detection of optimum bands for estimating species diversity within Shoolpaneshwar Wildlife Sanctuary (SWS), India. Narrow hyperspectral bands of EO-1 Hyperion image were screened and the best optimum wavelength from visible and Near Infrared (NIR) regions were identified based on coefficient of determination (r2) between band reflectance and in situ measured species diversity. For in situ species diversity measurements, quadrat sampling was carried out in SWS and different Diversity Indices (DIs) namely the Shannon Weiner DI, Margalef DI, McIntosh DI and Brillouin DI were calculated. The identified optimum wavelengths were then employed for modifying 38 existing spectral indices which were then investigated for testing their relation with the in situ DIs. The obtained optimum bands in visible and NIR regions were found to be in correspondence with four DIs. Among several indices used in this study, during validation, modified Non-linear index, modified Red Edge Position Index, modified Structure Insensitive Pigment Index and modified Red Green Ratio Index were identified as the best hyperspectral indices for determining Shannon Weiner DI, Margalef DI, McIntosh DI and Brillouin DI, respectively.
8 citations
References
More filters
TL;DR: The focus is on regression problems, which are those in which one of the measures, the dependent Variable, is of special interest, and the authors wish to explore its relationship with the other variables.
Abstract: Model fitting is an important part of all sciences that use quantitative measurements. Experimenters often explore the relationships between measures. Two subclasses of relationship problems are as follows: • Correlation problems: those in which we have a collection of measures, all of interest in their own right, and wish to see how and how strongly they are related. • Regression problems : those in which one of the measures, the dependent Variable, is of special interest, and we wish to explore its relationship with the other variables. These other variables may be called the independent Variables, the predictor Variables, or the coVariates. The dependent variable may be a continuous numeric measure such as a boiling point or a categorical measure such as a classification into mutagenic and nonmutagenic. We should emphasize that using the words ‘correlation problem’ and ‘regression problem’ is not meant to tie these problems to any particular statistical methodology. Having a ‘correlation problem’ does not limit us to conventional Pearson correlation coefficients. Log-linear models, for example, measure the relationship between categorical variables in multiway contingency tables. Similarly, multiple linear regression is a methodology useful for regression problems, but so also are nonlinear regression, neural nets, recursive partitioning and k-nearest neighbors, logistic regression, support vector machines and discriminant analysis, to mention a few. All of these methods aim to quantify the relationship between the predictors and the dependent variable. We will use the term ‘regression problem’ in this conceptual form and, when we want to specialize to multiple linear regression using ordinary least squares, will describe it as ‘OLS regression’. Our focus is on regression problems. We will use y as shorthand for the dependent variable and x for the collection of predictors available. There are two distinct primary settings in which we might want to do a regression study: • Prediction problems:We may want to make predictions of y for future cases where we know x but do not knowy. This for example is the problem faced with the Toxic Substances Control Act (TSCA) list. This list contains many tens of thousands of compounds, and there is a need to identify those on the list that are potentially harmful. Only a small fraction of the list however has any measured biological properties, but all of them can be characterized by chemical descriptors with relative ease. Using quantitative structure-activity relationships (QSARs) fitted to this small fraction to predict the toxicities of the much larger collection is a potentially cost-effective way to try to sort the TSCA compounds by their potential for harm. Later, we will use a data set for predicting the boiling point of a set of compounds on the TSCA list from some molecular descriptors. • Effect quantification:We may want to gain an understanding of how the predictors enter into the relationship that predicts y. We do not necessarily have candidate future unknowns that we want to predict, we simply want to know how each predictor drives the distribution of y. This is the setting seen in drug discovery, where the biological activity y of each in a collection of compounds is measured, along with molecular descriptors x. Finding out which descriptors x are associated with high and which with low biological activity leads to a recipe for new compounds which are high in the features associated positively with activity and low in those associated with inactivity or with adverse side effects. These two objectives are not always best served by the same approaches. ‘Feature selection’ skeeping those features associated withy and ignoring those not associated with y is very commonly a part of an analysis meant for effect quantification but is not necessarily helpful if the objective is prediction of future unknowns. For prediction, methods such as partial least squares (PLS) and ridge regression (RR) that retain all features but rein in their contributions are often found to be more effective than those relying on feature selection. What Is Overfitting? Occam’s Razor, or the principle of parsimony, calls for using models and procedures that contain all that is necessary for the modeling but nothing more. For example, if a regression model with 2 predictors is enough to explainy, then no more than these two predictors should be used. Going further, if the relationship can be captured by a linear function in these two predictors (which is described by 3 numbers sthe intercept and two slopes), then using a quadratic violates parsimony. Overfitting is the use of models or procedures that violate parsimonysthat is, that include more terms than are necessary or use more complicated approaches than are necessary. It is helpful to distinguish two types of overfitting: • Using a model that is more flexible than it needs to be. For example, a neural net is able to accommodate some curvilinear relationships and so is more flexible than a simple linear regression. But if it is used on a data set that conforms to the linear model, it will add a level of complexity without * Corresponding author e-mail: doug@stat.umn.edu. 1 J. Chem. Inf. Comput. Sci. 2004,44, 1-12
1,931 citations
TL;DR: In this paper, the authors investigated the utility of high spectral and spatial resolution imagery for the automated species-level classification of individual tree crowns (ITCs) in a tropical rain forest (TRF).
714 citations
TL;DR: In this paper, the impact of these upgrades on predictions for AVIRIS viewing scenarios is discussed for both clear and clouded skies; the CK approach provides refined predictions for nadir and near-nadir viewing.
711 citations
TL;DR: In this paper, a computer code (acronym 5S) is developed that allows estimation of the solar radiation backscattered by the Earth-surface-atmosphere system, as it is observed by a satellite sensor.
Abstract: A computer code (acronym 5S) has been developed that allows estimation of the solar radiation backscattered by the Earth-surface-atmosphere system, as it is observed by a satellite sensor. Given the Lambertian ground reflectance, the apparent reflectance of the observed pixel is estimated by taking into account the effects of gaseous absorption, scattering by molecules and aerosols and, to some extent, inhomogeneity in the ground reflectance. The input parameters (observation geometry, atmosphere model, ground reflectance and spectral band) can be either selected from some proposed standard conditions (e.g. spectral bands of a satellite sensor) or user-defined. Besides the pixel apparent reflectance, the code provides the gaseous transmittance, the irradiance at the surface and the different contributions to the satellite signal according to the origin of the measured radiance. Some complementary results are also available; among others, benchmark calculations permit assessment of the code accuracy.
615 citations
TL;DR: Study of the merit function in the numerical inversion showed that red edge optical indices used in the minimizing function such as R/sub 750//R/sub 710/ perform better than when all single spectral reflectance channels from hyper-spectral airborne CASI data are used, and in addition, the effect of shadows and LAI variation are minimized.
Abstract: Radiative transfer theory and modeling assumptions were applied at laboratory and field scales in order to study the link between leaf reflectance and transmittance and canopy hyper-spectral data for chlorophyll content estimation. This study was focused on 12 sites of Acer saccharum M. (sugar maple) in the Algoma Region, Canada, where field measurements, laboratory-simulation experiments, and hyper-spectral compact airborne spectrographic imager (CASI) imagery of 72 channels in the visible and near-infrared region and up to 1-m spatial resolution data were acquired in the 1997, 1998, and 1999 campaigns. A different set of 14 sites of the same species were used in 2000 for validation of methodologies. Infinite reflectance and canopy reflectance models were used to link leaf to canopy levels through radiative transfer simulation. The closed and dense (LAI>4) forest canopies of Acer saccharum M. used for this study, and the high spatial resolution reflectance data targeting crowns, allowed the use of optically thick simulation formulae and turbid-medium SAILH and MCRM canopy reflectance models for chlorophyll content estimation by scaling-up and by numerical model inversion approaches through coupling to the PROSPECT leaf radiative transfer model. Study of the merit function in the numerical inversion showed that red edge optical indices used in the minimizing function such as R/sub 750//R/sub 710/ perform better than when all single spectral reflectance channels from hyper-spectral airborne CASI data are used, and in addition, the effect of shadows and LAI variation are minimized.
603 citations