Hyperspectral Remote Sensing: Theory and Applications

Hyperspectral Remote Sensing
Contents
Section I Introduction to Hyperspectral Remote Sensing and Principles of Theory and Data Processing1
Section II Hyperspectral Remote Sensing Application in Vegetation93
Section III Hyperspectral Remote Sensing Application in Water, Snow, Urban Research165
Section IV Hyperspectral Remote Sensing Application in Soil and Mineral Exploration247
Section V Hyperspectral Remote Sensing: Multi-sensor, Fusion and Indices applications for Pollution Detection and Other App...
Section VI Hyperspectral Remote Sensing: Challenges, Future Pathway for Research & Emerging Applications427
Copyright
List of contributors
Biography
Foreword
Preface
1 Revisiting hyperspectral remote sensing: origin, processing, applications and way forward
1.1 Introduction
1.2 Origin of hyperspectral remote sensing
1.3 Atmospheric correction: a primary step in preprocessing hyperspectral images
1.4 Empirical and radiative transfer models
1.5 Applications of hyperspectral remote sensing
1.5.1 Vegetation analysis
1.5.2 Urban analysis
1.5.3 Mineral identification
1.5.4 Water quality
1.5.5 Agricultural applications
1.6 Way forward
Acknowledgment
References
2 Spectral smile correction for airborne imaging spectrometers
2.1 Introduction
2.2 Illumination effects on the spectral smile of airborne hyperspectral images
2.3 The modified trend line smile correction method
2.4 Implementation and results
2.4.1 Data
2.4.2 Implementation and results
2.5 Evaluation and discussion
2.6 Conclusion
List of abbreviations
References
3 Anomaly detection in hyperspectral remote sensing images
3.1 Introduction
3.1.1 An example of hyperspectral anomaly detection
3.1.2 Literature review of hyperspectral anomaly detection
3.2 Methods
3.2.1 Gaussian model: the RX detector
3.2.2 Extension of the model for local, partial, or multivariate Gaussian
3.2.3 Other approaches: non-Gaussian backgrounds
3.2.3.1 One-class support vector machine
3.2.3.2 Collaborative-based detector
3.2.3.3 Hidden Markov model-based detector
3.2.3.4 High-order two-dimensional crossing filter-based detector
3.3 Experiments
3.3.1 Detection datasets
3.3.1.1 HYDICE urban
3.3.1.2 HyMap Cooke City
3.3.1.3 AVIRIS San Diego
3.3.2 Experiment procedure
3.3.3 Results and discussion
3.4 Conclusion
Acknowledgment
List of abbreviations
List of symbols
References
4 Atmospheric parameter retrieval and correction using hyperspectral data
4.1 Introduction
4.2 Atmospheric correction techniques
4.2.1 Nonphysical models for atmospheric correction
4.2.2 Physics-based models for atmospheric correction
4.3 Aerosol retrieval method
4.4 Water vapor and other trace gas retrieval
4.5 Atmospheric correction results and discussion
4.6 Conclusion
List of abbreviations
List of symbols
References
5 Hyperspectral image classifications and feature selection
5.1 Introduction
5.2 Modified radial basis function neural network
5.3 Bayesian framework for feature selection
5.4 Study area and data sources
5.5 Results
5.6 Conclusion
List of abbreviations
List of symbols
Acknowledgment
References
6 Identification of functionally distinct plants using linear spectral mixture analysis
6.1 Introduction
6.2 Plant functional traits
6.3 Plant functional types
6.4 Remote sensing in identification of plant functional types or functionally distinct plants
6.5 Hyperspectral remote sensing in plant functional types identification
6.6 Spectral mixture analysis
6.7 Study site
6.8 Materials and methodology
6.8.1 Ground data collection and analysis
6.9 Satellite data and their analysis
6.10 Results and discussion
6.11 Conclusion
Acknowledgments
List of Abbreviation
References
7 Estimation of chengal trees relative abundance using coarse spatial resolution hyperspectral systems
7.1 Introduction
7.2 Materials and methodology
7.2.1 Study area
7.2.2 Hyperspectral system
7.2.3 Ancillary data (topographic map, census data, and spectral radiometer data)
7.2.3.1 Estimation of tree height from diameter at breast height
7.2.3.2 Estimation of tree crown from tree height
7.2.4 Spectral radiometer data collection
7.2.5 Hyperspectral image preprocessing
7.2.6 Canopy fractional cover
7.2.7 Vegetation index used in canopy fractional cover
7.2.8 Mixture tuned matched filtering
7.2.9 Relative abundance assessment
7.3 Results and analysis
7.3.1 Chengal trees relative abundance estimation using mixture tuned matched filtering
7.3.2 Relative abundance of chengal trees estimation by modified canopy fractional cover
7.4 Discussion
7.5 Conclusion
List of abbreviations
References
8 Hyperspectral remote sensing in precision agriculture: present status, challenges, and future trends
8.1 Introduction
8.2 Multispectral remote sensing in precision agriculture
8.2.1 Advantages
8.2.2 Multispectral data limitations in precision agriculture
8.2.3 Advantages of hyperspectral over multispectral data
8.2.4 Precision farming requirement
8.2.5 Spaceborne remote sensing: advantages and disadvantages
8.3 Hyperspectral sensors: present status
8.4 Hyperspectral data in agriculture
8.4.1 Recent approaches
8.4.1.1 Analytical spectral device field radio spectrometer
8.4.1.2 Global positioning system-guided unmanned aerial vehicles employing hyperspectral data
8.4.2 Case studies
8.4.2.1 Field spectroradiometry
8.4.2.2 Crop characterization and discrimination
8.4.2.3 Land Use Land Cover (LULC) mapping
8.4.2.4 Insect, invasive plant species, and plant disease monitoring
8.4.2.5 Drought mapping
8.5 Hyperspectral sensors: future missions
8.6 Conclusion
Acknowledgments
List of abbreviations
References
9 Discriminating tropical grasses grown under different nitrogen fertilizer regimes in KwaZulu-Natal, South Africa
9.1 Introduction
9.2 Materials and methods
9.2.1 Study area
9.2.2 Experimental design
9.2.3 Field data collection and laboratory analyses
9.2.4 Statistical data analyses
9.2.4.1 Partial least squares classification ensembles
9.2.4.2 Variable importance in the projection
9.2.4.3 Accuracy assessment
9.3 Results
9.3.1 The classification of fertilizer treatments across different phenological stages
9.3.2 Comparing the performance of PLS-DA and PLS-LDA after optimization
9.4 Discussion
9.4.1 Discrimination of different nitrogen fertilizer treatment regimes and characterization of the soil–plant nitrogen rel...
9.4.2 Comparing the performance of PLS-DA and PLS-LDA classification ensembles
9.5 Conclusion
Acknowledgment
Author contributions
Funding
Conflict of Interest
List of abbreviations
List of symbols
References
10 Effect of contamination and adjacency factors on snow using spectroradiometer and hyperspectral images
10.1 Remote sensing of snow
10.2 Snow spectra
10.3 Hyperspectral remote sensing
10.4 The experimental sites
10.5 Data used
10.6 Methodology used
10.6.1 Preprocessing of hyperion data
10.6.2 Snow grain size measurement
10.7 Contamination in snow
10.7.1 Soil contamination in snow
10.7.2 Coal contamination in snow
10.7.3 Carbon soot contamination in snow
10.7.4 Ash contamination in snow
10.7.5 Sparse vegetation in snow
10.7.6 Dust contamination in snow
10.7.7 Contamination of algae in snow
10.7.8 Contamination of sparse debris
10.7.9 Influence of mixed and contaminated snow on normalized differential snow index
10.7.10 Contamination index for different levels of contamination using spectroradiometer
10.8 Adjacent objects and their effects on snow reflectance
10.8.1 Adjacency effects
10.8.2 Liquid water content
10.8.3 Clear waterbody
10.8.4 Vegetation
10.9 Spectral unmixing methods for satellite image classification
10.9.1 Linear unmixing model
10.9.2 Nonlinear unmixing model
10.10 Conclusion
List of abbreviations
List of symbols
References
11 Remote sensing of inland water quality: a hyperspectral perspective
11.1 Introduction
11.2 Hyperspectral remote sensing
11.3 Methodology: field and satellite measurements
11.3.1 In situ hyperspectral radiometry
11.3.2 Water sample laboratory analysis
11.3.2.1 Chlorophyll-a
11.3.2.2 Colored dissolved organic matter
11.3.2.3 Total suspended matter
11.3.3 Atmospheric correction of inland waters
11.3.4 Retrieval of water quality parameters
11.3.4.1 Empirical relations
11.3.4.2 Semianalytical solutions
11.3.4.3 Software
11.4 Interpretation of the spectral signatures
11.4.1 Chlorophyll-a: the fundamental measure of phytoplankton biomass
11.4.2 Colored dissolved organic matter
11.4.3 Sediment laden and clear river water
11.4.4 Cyanobacterial bloom
11.4.5 Aquatic macrophytes
11.5 Conclusion
Acknowledgments
List of abbreviations
List of symbols
References
12 Efficacy of hyperspectral data for monitoring and assessment of wetland ecosystem
12.1 Introduction
12.1.1 Wetland ecosystem and its services
12.1.2 Role of Ramsar Convention in wetland ecosystems
12.1.3 Global status of wetland ecosystem
12.1.4 Indian status of wetland ecosystems
12.2 Monitoring and assessment of wetlands with multispectral remote sensing
12.3 Monitoring and assessment of wetlands with hyperspectral remote sensing
12.4 Details of hyperspectral images for wetland monitoring
12.4.1 Airborne hyperspectral sensors
12.4.2 Spaceborne hyperspectral sensors
12.5 Application of hyperspectral images for wetland ecosystems
12.6 Future scope and challenges of hyperspectral remote sensing for wetland ecosystem
12.6.1 Future scopes
12.6.2 Challenges
12.7 Applicability of hyperion image for Sambhar Salt Lake, a saline wetland
12.7.1 Study area
12.7.2 Methodology
12.7.3 Results
12.7.4 Discussion
Acknowledgments
List of abbreviations
List of symbols
References
13 Spectroradiometry as a tool for monitoring soil contamination by heavy metals in a floodplain site
13.1 Introduction
13.2 Distribution and vulnerabilities of heavy metals in the United Kingdom
13.3 Materials and methods
13.3.1 Study area and soil sampling
13.3.2 Field and laboratory spectral measurements
13.3.3 Geochemical analysis of the soil samples
13.3.4 Data processing and statistics
13.4 Results and discussion
13.4.1 Soil descriptive statistics
13.4.2 Creation of field- and lab-based spectral libraries
13.4.3 Statistical discrimination analysis
13.4.4 Model development and validation
13.5 Conclusion
Acknowledgments
List of abbreviations
List of symbols
References
Further reading
14 Hyperspectral remote sensing applications in soil: a review
14.1 Introduction
14.2 Hyperspectral remote sensing application in soil mineral identification
14.3 Hyperspectral remote sensing application in soil nutrient prediction
14.4 Hyperspectral remote sensing application in soil organic carbon estimation
14.5 Hyperspectral remote sensing application in soil moisture retrieval
14.6 Hyperspectral remote sensing application in soil salinity detection
14.7 Hyperspectral remote sensing application in soil texture acquisition
14.8 Opportunities and challenges
Acknowledgments
Conflict of interest
List of abbreviations
References
15 Mineral exploration using hyperspectral data
15.1 Introduction
15.2 Spectroscopy of rocks and minerals
15.2.1 Olivine
15.2.2 Pyroxene
15.2.3 Carbonate minerals
15.2.4 Clay and mica minerals
15.2.5 Iron oxides and iron hydroxides
15.2.6 Sulfide
15.3 Hyperspectral sensors suitable for mineral exploration
15.4 Broad overview of hyperspectral data processing steps for geological exploration
15.4.1 Signal to noise ratio estimation
15.4.2 Data dimensionality reduction
15.4.3 Endmember extraction
15.4.4 Spectral mapping
15.5 Application of hyperspectral remote sensing in mineral exploration
15.6 Requirement and future research focus
15.6.1 Data requirement and related approach
15.6.2 Requirement of improving data processing approach
List of abbreviations
References
Further reading
16 Metrological hyperspectral image analysis through spectral differences
16.1 Introduction
16.2 Is metrology justified for remote sensing?
16.3 Towards a metrological spectral difference function
16.3.1 Expected theoretical properties
16.3.2 Metrological validation of monotonicity property
16.3.3 Metrological validation of discrimination performances
16.3.4 Demonstration of use in oil detection
16.4 Assessing the nonuniformity of the spectral world
16.4.1 Statistics on spectral differences
16.4.2 Histogram of spectral differences
16.4.3 Variance–covariance in a spectral difference space
16.5 Application in forest mapping
16.5.1 Measuring variability within a known dataset
16.5.2 Mapping tree genus
16.6 Conclusion
List of abbreviations
List of symbols
References
17 Improving the detection of cocoa bean fermentation-related changes using image fusion
17.1 Introduction
17.2 Methodology
17.2.1 Image acquisition
17.2.2 Image preprocessing
17.2.3 Image fusion
17.2.4 Morphological feature data
17.3 Experiments
17.3.1 Bean coat
17.3.2 Bean-cuts
17.4 Discussion
Acknowledgments
Abbreviations
Symbols
References
18 Noninvasive detection of plant parasitic nematodes using hyperspectral and other remote sensing systems
18.1 Introduction to noninvasive detection
18.2 Introduction to plant parasitic nematodes
18.2.1 Plant parasitic nematodes
18.2.2 Cyst and root-knot nematodes
18.2.3 Pinewood nematode
18.3 Examples of noninvasive detection of plant parasitic nematodes
18.3.1 Characteristics of nematode infestations
18.3.2 Remote sensing of nematode infestations
18.3.3 Hyperspectral remote sensing of nematode infestations
18.3.4 Remote sensing of cyst, root-knot, and pinewood nematodes
18.4 Patents in the field of remote sensing of nematode infestations
18.5 Conclusion
Acknowledgments
List of abbreviation
References
19 Evaluating the performance of vegetation indices for detecting oil pollution effects on vegetation using hyperspectral (...
19.1 Introduction
19.2 Materials and methods
19.2.1 Study area
19.2.2 Establishment of study transects
19.2.3 Field survey of vascular plant species
19.2.4 Soil sampling and analysis
19.2.5 Data acquisition and preprocessing
19.2.5.1 Hyperspectral (Hyperion) image
19.2.5.2 Multispectral (Sentinel-2A) image
19.2.6 Image to image registration
19.2.7 Gram–Schmidt fusion of multispectral and hyperspectral images
19.2.8 Vegetation indices
19.2.9 Vegetation response to oil pollution
19.2.10 Statistical analysis
19.3 Results
19.3.1 Total petroleum hydrocarbon concentration in soil
19.3.2 Registration of images
19.3.3 Fusion of multispectral and hyperspectral images
19.3.4 Effect of oil pollution on vegetation indices
19.3.5 Response of vegetation parameters to oil pollution
19.3.6 Detecting vegetation response to oil pollution using vegetation indices
19.4 Discussion
19.4.1 Changes in chlorophyll concentration of polluted vegetation
19.4.2 Changes in species richness of polluted vegetation
19.4.3 Changes in vegetation abundance on polluted transects
19.5 Conclusion
Acknowledgments
References
20 Hyperspectral vegetation indices to detect hydrocarbon pollution
20.1 Introduction
20.1.1 Hydrocarbons and vegetation interactions
20.1.2 Vegetation indices for vegetation stress caused by hydrocarbons
20.2 Materials and methods
20.2.1 Study area
20.2.2 Sampling process
20.2.3 Chlorophyll meter readings
20.2.4 Foliar biophysical and biochemical measurements
20.2.4.1 Foliar water content
20.2.4.2 Dry matter content
20.2.5 Spectroradiometer measurements
20.2.6 PROSPECT model
20.2.7 Hyperion EO-1 hyperspectral satellite image: data acquisition and preprocessing
20.2.8 MTCI vegetation index
20.2.9 Vegetation indices suitable to detect hydrocarbon pollution
20.2.10 Exploring new vegetation indices for detecting hydrocarbon pollution in the tropical rainforest
20.2.11 Results
20.2.11.1 Exploring new vegetation indices for vegetation affected by petroleum pollution
20.3 Discussion
20.3.1 Exploring new vegetation indices for vegetation affected by petroleum pollution
20.4 Conclusion
List of abbreviations
List of symbols
References
21 Future perspectives and challenges in hyperspectral remote sensing
21.1 Introduction
21.2 Challenges of hyperspectral imaging systems
21.3 Conclusion
Acknowledgments
List of abbreviations
References
Author Index
Subject Index