Towards tree-level evapotranspiration estimation with small UAVs in precision agriculture / / Haoyu Niu and YangQuan Chen

• Record Nr.: UNINA9910624384403321
• Autore: Niu Haoyu
• Pubbl/distr/stampa: Cham, Switzerland : , : Springer, , [2022]; ©2022
• ISBN: 3-031-14937-8
• Descrizione fisica: 1 online resource (176 pages)
• Disciplina: 631
• Soggetti: Precision farming; Evapotranspiration - Measurement; Drone aircraft in remote sensing
• Lingua di pubblicazione: Inglese
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Nota di bibliografia: Includes bibliographical references and index.
• Nota di contenuto: Intro -- Preface -- Acknowledgments -- Contents -- Acronyms -- List of Figures -- List of Tables -- 1 Introduction -- 1.1 What Is Evapotranspiration Estimation? -- 1.2 Challenges and Opportunities -- 1.3 Smart Big Data in Precision Agriculture: Acquisition and Advanced Analytics -- 1.3.1 What Is Smart Big Data in Precision Agriculture? -- 1.3.2 Plant Physiology-Informed Machine Learning: A New Frontier for Precision Agriculture -- 1.3.3 Big Data Acquisition and Advanced Analytics -- 1.3.4 Fractional Calculus (FC) and Fractional-Order Thinking (FOT) -- 1.3.5 Complexity and Inverse Power Laws (IPLs) -- 1.3.6 Heavy-Tailed Distributions -- 1.3.6.1 The Lévy Distribution -- 1.3.6.2 The Mittag-Leffler PDF -- 1.3.6.3 The Weibull Distribution -- 1.3.6.4 The Cauchy Distribution -- 1.3.6.5 The Pareto Distribution -- 1.3.6.6 The α-Stable Distribution -- 1.3.6.7 Mixture Distributions -- 1.3.6.8 The Gaussian Distribution -- 1.3.6.9 The Laplace Distribution -- 1.3.7 Big Data, Variability, and FC -- 1.3.7.1 The Hurst Parameter, fGn, and fBm -- 1.3.7.2 Fractional Lower-Order Moments (FLOMs) -- 1.3.7.3 Fractional Autoregressive Integrated Moving Average (FARIMA) and Gegenbauer Autoregressive Moving Average (GARMA) -- 1.3.7.4 Continuous-Time Random Walk (CTRW) -- 1.3.7.5 Unmanned Aerial Vehicles (UAVs) and Precision Agriculture -- 1.3.8 Optimal Machine Learning and Optimal Randomness -- 1.3.8.1 Derivative-Free Methods ; -- 1.3.8.2 Gradient-Based Methods -- 1.3.8.3 The Nesterov Accelerated Gradient Descent (NAGD) -- 1.4 Main Contributions -- 1.5 Book Organization -- 1.6 Results Reproducibility -- References -- 2 Small Unmanned Aerial Vehicles (UAVs) and Remote Sensing Payloads -- 2.1 The UAV Platform -- 2.2 Lightweight Sensors -- 2.2.1 RGB Camera -- 2.2.2 Multispectral Camera -- 2.2.3 Shortwave Infrared Camera -- 2.2.4 Thermal Camera.; 2.3 UAV Image Acquisition and Processing -- 2.3.1 Flight Mission Design -- 2.3.2 UAV Image Processing -- 2.4 Challenges and Opportunities -- 2.4.1 UAVs -- 2.4.2 UAV Path Planning and Image Processing -- 2.4.3 Preflight Path Planning -- 2.4.4 Multispectral Image Calibration -- 2.4.5 Thermal Camera Calibration and Image Processing -- 2.4.6 Image Stitching and Orthomosaick Image Generation -- 2.5 Case Study I: A UAV Resolution and Waveband Aware Path Planning for Irrigation -- 2.5.1 Introduction -- 2.5.2 Material and Methods -- 2.5.2.1 Onion Study Area -- 2.5.2.2 A UAV Platform and Sensors -- 2.5.2.3 UAV Image Collection and Preprocessing -- 2.5.2.4 Principal Component Analysis -- 2.5.2.5 Linear Discriminant Analysis -- 2.5.3 Results and Discussion -- 2.5.3.1 UAV Flight Height or Resolution's Effect -- 2.5.3.2 Wavebands Configuration's Effect -- 2.5.4 Conclusions -- 2.6 Case Study II: A Detailed Study on Accuracy of Uncooled Thermal Cameras -- 2.6.1 Introduction -- 2.6.2 Material and Methods -- 2.6.2.1 Study Site -- 2.6.2.2 Image Collection -- 2.6.2.3 Groundtruth Data Collection -- 2.6.3 Results and Discussion -- 2.6.3.1 Experiment Setup -- 2.6.3.2 Thermal Camera Warm-Up Time -- 2.6.3.3 Calibration Experiment -- 2.6.3.4 The View Angle of Thermal Camera -- 2.6.3.5 The Effect of Stitching -- 2.6.4 Conclusions and Future Work -- 2.7 Case Study III: High Spatial Resolution Has Little Impact on NDVI Mean Value -- 2.7.1 Introduction -- 2.7.2 Material and Methods -- 2.7.2.1 The Study Site -- 2.7.2.2 The UAV and the Multispectral Sensor -- 2.7.2.3 Details of the UAV Imagery Dataset -- 2.7.3 Results and Discussion -- 2.7.3.1 The Relationship Between NDVI and UAV Flight Height -- 2.7.3.2 Individual Tree Canopy Segmentation Using Support Vector Machine (SVM) -- 2.7.3.3 Entropy of Individual Tree-Level NDVI Image -- 2.7.4 Conclusions and Future Work.; 2.8 Chapter Summary -- References -- 3 ET Estimation Methods with Small UAVs: A Literature Review -- 3.1 Introduction -- 3.2 Related Work -- 3.2.1 One-Source Energy Balance (OSEB) -- 3.2.2 High-Resolution Mapping of ET (HRMET) -- 3.2.3 Machine Learning (ML) and Artificial Neural Networks (ANN) -- 3.2.4 Two-Source Energy Balance (TSEB) Models -- 3.2.5 Dual-Temperature-Difference (DTD) Model -- 3.2.6 Surface Energy Balance Algorithm for Land (SEBAL) -- 3.2.7 Mapping Evapotranspiration at High Resolution with Internalized Calibration (METRIC) -- 3.3 Existing ET Estimation Methods with UAVs: Results and Discussion -- 3.3.1 OSEB and TSEB Models -- 3.3.2 HARMET Model -- 3.3.3 Machine Learning and Neural Networks -- 3.3.4 TSEB and DTD Models -- 3.3.5 TSEB and DATTUTDUT Models -- 3.3.6 SEBAL Model -- 3.3.7 METRIC and METRIC-HR Models -- 3.4 Chapter Summary -- References -- 4 Estimating ET Using Stochastic Configuration Network and UAV-Based Crop Coefficients -- 4.1 Introduction -- 4.2 Material and Methods -- 4.2.1 Pomegranate Study Area -- 4.2.2 The UAV Platform and Multispectral Camera -- 4.2.3 UAV Image Collection and Preprocessing -- 4.2.4 Deep Stochastic Configuration Networks (DeepSCNs) -- 4.3 Results and Discussion -- 4.3.1 Seasonal Kc and NDVI -- 4.3.2 Regression Models for Kc and NDVI -- 4.4 Conclusions -- 4.5 Case Study: Optimal Randomness for SCN with Heavy-Tailed Distributions -- 4.5.1 Introduction -- 4.5.2 SCN with Heavy-Tailed PDFs -- 4.5.3 A Regression Model and Parameter ; Tuning -- 4.5.3.1 Performance Comparison Among SCNs with Heavy-Tailed PDFs -- 4.5.4 MNIST Handwritten Digit Classification -- 4.5.4.1 Performance Comparison Among SCNs on MNIST -- 4.6 Chapter Summary -- References -- 5 Reliable Tree-Level ET Estimation Using Lysimeter and UAV Multispectral Imagery -- 5.1 Introduction -- 5.2 Material and Methods -- 5.2.1 Study Site Description.; 5.2.2 UAV Image Collection and Processing -- 5.3 Results and Discussion -- 5.3.1 Determination of Individual Tree Kc from NDVI -- 5.3.2 The Spatial Variability Mapping of Kc and ETc -- 5.3.3 Performance of the Individual Tree-Level ET Estimation -- 5.3.4 Conclusion -- 5.4 Chapter Summary -- References -- 6 Tree-Level Water Status Inference Using UAV Thermal Imagery and Machine Learning -- 6.1 Introduction -- 6.2 Material and Methods -- 6.2.1 Experimental Site and Irrigation Management -- 6.2.2 Ground Truth: Infrared Canopy and Air Temperature -- 6.2.3 Thermal Infrared Remote Sensing Data -- 6.2.3.1 UAV Thermal Image Collection and Processing -- 6.2.3.2 Tree Canopy Segmentation Using Support Vector Machine (SVM) -- 6.2.4 Complexity-Informed Machine Learning (CIML) -- 6.2.5 Principle of Tail Matching -- 6.2.5.1 Pareto Distribution -- 6.2.6 Machine Learning Classification Algorithms -- 6.2.7 Image Preprocessing for the CNN Model -- 6.3 Results and Discussion -- 6.3.1 Comparison of Canopy Temperature Per Tree Based on Ground Truth and UAV Thermal Imagery -- 6.3.2 The Relationship Between ΔT and Irrigation Treatment -- 6.3.3 The Classification Performance of CIML on Irrigation Treatment Levels -- 6.3.4 The Performance of the CNN Model -- 6.4 Conclusion and Future Research -- 6.5 Chapter Summary -- References -- 7 Conclusion and Future Research -- 7.1 Conclusions -- 7.2 Future Research -- References -- Index.