Precision Soil Moisture Monitoring Through Drone-Based Hyperspectral Imaging and PCA-Driven Machine Learning

Sensors, Journal Year: 2025, Volume and Issue: 25(3), P. 782 - 782

Published: Jan. 28, 2025

Accurately estimating soil moisture at multiple depths is essential for sustainable farming practices, as it supports efficient irrigation management, optimizes crop yields, and conserves water resources. This study integrates a drone-mounted hyperspectral sensor with machine learning techniques to enhance estimation 10 cm 30 in cornfield. The primary aim was understand the relationship between root zone content canopy reflectance, pinpoint where this most significant, identify informative wavelengths, train model using those wavelengths estimate moisture. Our results demonstrate that PCA effectively detected critical variables estimation, ANN outperforming other algorithms, including Random Forest (RF), Support Vector Regression (SVR), Gradient Boosting (XGBoost). Model comparisons irrigated non-irrigated treatments showed plots could be estimated greater accuracy across various dates. finding indicates plants experiencing high stress exhibit more significant spectral variability their canopy, enhancing correlation zone. Moreover, over growing season, when corn exhibits chlorophyll increased resilience environmental stressors, spectrum weakens. Error analysis revealed lowest relative errors depth, aligning periods of elevated shallower levels, which drove deeper growth strengthened reflectance relationship. corresponded lower RMSE values, highlighting improved accuracy.

Language: Английский

---

Optimal band selection and transfer in drone-based hyperspectral images for plant-level vegetable crops identification using statistical-swarm intelligence (SSI) hybrid algorithms

Ecological Informatics, Journal Year: 2025, Volume and Issue: unknown, P. 103051 - 103051

Published: Jan. 1, 2025

Language: Английский

---

Urban spatial structure and air quality in the United States: Evidence from a longitudinal approach

Environment International, Journal Year: 2024, Volume and Issue: 190, P. 108871 - 108871

Published: July 3, 2024

Previous studies on the relationship between urban form and air quality: (1) report mixed results among specific aspects of spatial structure (e.g., expansion, form, or shape) (2) use primarily cross-sectional approaches with a single year data. This study takes advantage multi-decade, longitudinal approach to investigate impact population-weighted concentrations PM

Language: Английский

---

Pasture Biomass Estimation Using Ultra-High-Resolution RGB UAVs Images and Deep Learning

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(24), P. 5714 - 5714

Published: Dec. 13, 2023

The continuous assessment of grassland biomass during the growth season plays a vital role in making informed, location-specific management choices. implementation precision agriculture techniques can facilitate and enhance these decision-making processes. Nonetheless, depends on availability prompt precise data pertaining to plant characteristics, necessitating both high spatial temporal resolutions. Utilizing structural spectral attributes extracted from low-cost sensors unmanned aerial vehicles (UAVs) presents promising non-invasive method evaluate traits, including above-ground height. Therefore, main objective was develop an artificial neural network capable estimating pasture by using UAV RGB images canopy height models (CHM) growing over three common types paddocks: Rest, bale grazing, sacrifice. Subsequently, this study first explored variation color-related features derived statistics CHM image values under different levels growth. Then, ANN model trained for accurate volume estimation based rigorous employing statistical criteria ground observations. demonstrated level precision, yielding coefficient determination (R2) 0.94 root mean square error (RMSE) 62 (g/m2). evaluation underscores critical ultra-high-resolution photogrammetric CHMs red, green, blue (RGB) capturing meaningful variations enhancing model’s accuracy across diverse paddock types, rest, sacrifice paddocks. Furthermore, sensitivity areas with minimal or virtually absent period is visually generated maps. Notably, it effectively discerned low-biomass regions grazing paddocks reduced impact compared other types. These findings highlight versatility range scenarios, well suited deployment various environmental conditions.

Language: Английский

---

Enhancing Hyper-Spectral Image Classification with Reinforcement Learning and Advanced Multi-Objective Binary Grey Wolf Optimization

Computers, materials & continua/Computers, materials & continua (Print), Journal Year: 2024, Volume and Issue: 79(3), P. 3469 - 3493

Published: Jan. 1, 2024

Hyperspectral (HS) image classification plays a crucial role in numerous areas including remote sensing (RS), agriculture, and the monitoring of environment.Optimal band selection HS images is for improving efficiency accuracy classification.This process involves selecting most informative spectral bands, which leads to reduction data volume.Focusing on these key bands also enhances algorithms, as redundant or irrelevant can introduce noise lower model performance, are excluded.In this paper, we propose an approach using deep Q learning (DQL) novel multi-objective binary grey wolf optimizer (MOBGWO).We investigate MOBGWO optimal further enhance classification.In suggested MOBGWO, new sigmoid function introduced transfer modify wolves' position.The primary objective reduce number while maximizing accuracy.To evaluate effectiveness our approach, conducted experiments publicly available datasets, Pavia University, Washington Mall, Indian Pines datasets.We compared performance proposed method with several state-of-the-art (DL) machine (ML) long short-term memory (LSTM), neural network (DNN), recurrent (RNN), support vector (SVM), random forest (RF).Our experimental results demonstrate that Hybrid MOBGWO-DQL significantly improves traditional optimization DL techniques.MOBGWO-DQL shows greater classifying categories both datasets used.For Pine dataset, architecture achieved kappa coefficient (KC) 97.68% overall (OA) 94.32%.This was accompanied by lowest root mean square error (RMSE) 0.94, indicating very precise predictions minimal error.In case University demonstrated outstanding highest KC 98.72% impressive OA 96.01%.It recorded RMSE at 0.63, reinforcing its predictions.The clearly not only reaches highly accurate more quickly but maintains superior throughout training process.

Language: Английский

---

Binary Peacock Algorithm: A Novel Metaheuristic Approach for Feature Selection

Journal of Classification, Journal Year: 2024, Volume and Issue: 41(2), P. 216 - 244

Published: March 4, 2024

Language: Английский

---

Predicting PM10 Concentrations Using Evolutionary Deep Neural Network and Satellite-Derived Aerosol Optical Depth

Mathematics, Journal Year: 2023, Volume and Issue: 11(19), P. 4145 - 4145

Published: Sept. 30, 2023

Predicting particulate matter with a diameter of 10 μm (PM10) is crucial due to its impact on human health and the environment. Today, aerosol optical depth (AOD) offers high resolution wide coverage, making it viable way estimate PM concentrations. Recent years have also witnessed in-creasing promise in refining air quality predictions via deep neural network (DNN) models, out-performing other techniques. However, learning weights biases DNN task classified as an NP-hard problem. Current approaches such gradient-based methods exhibit significant limitations, risk becoming ensnared local minimal within multi-objective loss functions, substantial computational requirements, requirement for continuous objective functions. To tackle these challenges, this paper introduces novel approach that combines binary gray wolf optimizer (BGWO) improve optimization models pollution prediction. The BGWO algorithm, inspired by behavior wolves, used optimize both weight bias DNN. In proposed BGWO, sigmoid function transfer adjust position wolves. This study gathers meteorological data, topographic information, PM10 satellite images. Data preparation includes tasks noise removal handling missing data. evaluated through cross-validation using metrics correlation rate, R square, root-mean-square error (RMSE), accuracy. effectiveness BGWO-DNN framework compared seven machine (ML) models. experimental evaluation method data shows superior performance traditional ML BGWO-DNN, CapSA-DNN, BBO-DNN achieved lowest RMSE values 16.28, 19.26, 20.74, respectively. Conversely, SVM-Linear GBM algorithms displayed highest levels error, yielding 36.82 32.50, algorithm secured R2 (88.21%) accuracy (93.17%) values, signifying Additionally, between predicted actual model surpasses observes relatively stable during spring summer, contrasting fluctuations autumn winter.

Language: Английский

---

Intelligent Data Mining of Hyper Spectral Images for Feature Extraction

Published: March 15, 2024

This technical abstract describes the software of wise information mining strategies for analysis hyper-spectral imaging purpose extracting underlying features. Specifically, usage supervised and unsupervised device studying algorithms is explored, which include clustering, choice trees, help vector machines artificial neural networks. The techniques are in comparison on way to optimize consequences diverse consequences, including endmember extraction target detection. A singular technique was additionally proposed enhancing class accuracy lowering fake detections through use winger correction. paper outcomes reveal capability smart picture evaluation quite a number applications.

Language: Английский

---

Optimizing Soil Moisture Analysis with Drone-Based Hyperspectral Data and Pca-Enhanced Machine Learning

Published: Jan. 1, 2024

Language: Английский

---

A new feature selection algorithm combining genetic algorithm, exponential decay function, and machine learning to realize hyperspectral estimation of winter wheat leaf area index

Computers and Electronics in Agriculture, Journal Year: 2024, Volume and Issue: 230, P. 109851 - 109851

Published: Dec. 26, 2024

Language: Английский

---