Deep learning based computer vision approaches for smart agricultural applications

Artificial Intelligence in Agriculture, Journal Year: 2022, Volume and Issue: 6, P. 211 - 229

Published: Jan. 1, 2022

The agriculture industry is undergoing a rapid digital transformation and growing powerful by the pillars of cutting-edge approaches like artificial intelligence allied technologies. At core intelligence, deep learning-based computer vision enables various activities to be performed automatically with utmost precision enabling smart into reality. Computer techniques, in conjunction high-quality image acquisition using remote cameras, enable non-contact efficient technology-driven solutions agriculture. This review contributes providing state-of-the-art technologies based on learning that can assist farmers operations starting from land preparation harvesting operations. Recent works area were analyzed this paper categorized (a) seed quality analysis, (b) soil (c) irrigation water management, (d) plant health (e) weed management (f) livestock (g) yield estimation. also discusses recent trends such as generative adversarial networks (GAN), transformers (ViT) other popular architectures. Additionally, study pinpoints challenges implementing farmer’s field real-time. overall finding indicates convolutional neural are corner stone modern their architectures provide across terms accuracy. However, success approach lies building model dataset real-time solutions.

Language: Английский

---

An Integrated Statistical-Machine Learning Approach for Runoff Prediction

Sustainability, Journal Year: 2022, Volume and Issue: 14(13), P. 8209 - 8209

Published: July 5, 2022

Nowadays, great attention has been attributed to the study of runoff and its fluctuation over space time. There is a crucial need for good soil water management system overcome challenges scarcity other natural adverse events like floods landslides, among others. Rainfall–runoff (R-R) modeling an appropriate approach prediction, making it possible take preventive measures avoid damage caused by hazards such as floods. In present study, several data-driven models, namely, multiple linear regression (MLR), adaptive splines (MARS), support vector machine (SVM), random forest (RF), were used rainfall–runoff prediction Gola watershed, located in south-eastern part Uttarakhand. The model analysis was conducted using daily rainfall data 12 years (2009 2020) watershed. first 80% complete train model, remaining 20% testing period. performance models evaluated based on coefficient determination (R2), root mean square error (RMSE), Nash–Sutcliffe efficiency (NSE), percent bias (PBAIS) indices. addition numerical comparison, evaluated. Their performances graphical plotting, i.e., time-series line diagram, scatter plot, violin relative Taylor diagram (TD). comparison results revealed that four heuristic methods gave higher accuracy than MLR model. Among learning RF (RMSE (m3/s), R2, NSE, PBIAS (%) = 6.31, 0.96, 0.94, −0.20 during training period, respectively, 5.53, 0.95, 0.92, respectively) surpassed MARS, SVM, forecasting all cases studied. outperformed models’ periods. It can be summarized best-in-class delivers strong potential

Language: Английский

---

Prediction of meteorological drought and standardized precipitation index based on the random forest (RF), random tree (RT), and Gaussian process regression (GPR) models

Environmental Science and Pollution Research, Journal Year: 2023, Volume and Issue: 30(15), P. 43183 - 43202

Published: Jan. 17, 2023

Language: Английский

---

Comparative Assessment of Improved SVM Method under Different Kernel Functions for Predicting Multi-scale Drought Index

Water Resources Management, Journal Year: 2023, Volume and Issue: 37(3), P. 1367 - 1399

Published: Feb. 1, 2023

Language: Английский

---

Suspended sediment load prediction using sparrow search algorithm-based support vector machine model

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: June 5, 2024

Abstract Prediction of suspended sediment load (SSL) in streams is significant hydrological modeling and water resources engineering. Development a consistent accurate prediction model highly necessary due to its difficulty complexity practice because transportation vastly non-linear governed by several variables like rainfall, strength flow, supply. Artificial intelligence (AI) approaches have become prevalent resource engineering solve multifaceted problems modelling. The present work proposes robust incorporating support vector machine with novel sparrow search algorithm (SVM-SSA) compute SSL Tilga, Jenapur, Jaraikela Gomlai stations Brahmani river basin, Odisha State, India. Five different scenarios are considered for development. Performance assessment developed analyzed on basis mean absolute error (MAE), root squared (RMSE), determination coefficient (R 2 ), Nash–Sutcliffe efficiency (E NS ). outcomes SVM-SSA compared three hybrid models, namely SVM-BOA (Butterfly optimization algorithm), SVM-GOA (Grasshopper SVM-BA (Bat benchmark SVM model. findings revealed that successfully estimates high accuracy scenario V (3-month lag) discharge (current time-step 3-month as input than other alternatives RMSE = 15.5287, MAE 15.3926, E 0.96481. conventional performed the worst prediction. Findings this investigation tend claim suitability employed approach rivers precisely reliably. guarantees precision forecasted while significantly decreasing computing time expenditure, satisfies demands realistic applications.

Language: Английский

---

Comparative assessment of empirical and hybrid machine learning models for estimating daily reference evapotranspiration in sub-humid and semi-arid climates

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 20, 2025

Language: Английский

---

Drought indicator analysis and forecasting using data driven models: case study in Jaisalmer, India

Stochastic Environmental Research and Risk Assessment, Journal Year: 2022, Volume and Issue: 37(1), P. 113 - 131

Published: Aug. 1, 2022

Language: Английский

---

Data intelligence and hybrid metaheuristic algorithms-based estimation of reference evapotranspiration

Applied Water Science, Journal Year: 2022, Volume and Issue: 12(7)

Published: May 6, 2022

Abstract For developing countries, scarcity of climatic data is the biggest challenge, and model development with limited meteorological input critical importance. In this study, five intelligent hybrid metaheuristic machine learning algorithms, namely additive regression (AR), AR-bagging, AR-random subspace (AR-RSS), AR-M5P, AR-REPTree, were applied to predict monthly mean daily reference evapotranspiration (ET 0 ). purpose, two stations located in semi-arid region Pakistan used from period 1987 2016. The dataset includes maximum minimum temperature ( T max , min ), average relative humidity (RH avg wind speed U x sunshine hours n Sensitivity analysis through methods was determine effective parameters for ET modeling. results performed on all proved that RH Avg identified as most influential at studied station. From results, it revealed selected models predicted both greater precision. AR-REPTree furthest AR-M5P nearest observed point based performing indices stations. study concluded under aforementioned methodological framework, can yield higher accuracy predicting values, compared other algorithms.

Language: Английский

---

Assessment of Climate Change Impact on Snowmelt Runoff in Himalayan Region

Sustainability, Journal Year: 2022, Volume and Issue: 14(3), P. 1150 - 1150

Published: Jan. 20, 2022

Under different climate change scenarios, the current study was planned to simulate runoff due snowmelt in Lidder River catchment Himalayan region. A basic degree-day model, Snowmelt-Runoff Model (SRM), utilized assess hydrological consequences of climate. The performance SRM model during calibration and validation assessed using volume difference (Dv) coefficient determination (R2). Dv found be 11.7, −10.1, −11.8, 1.96, 8.6 2009–2014, respectively, while respective R2 0.96, 0.92, 0.95, 0.90, 0.94. values indicate that simulated closely agrees with observed values. findings were under three scenarios: (a) an increase precipitation by +20%, (b) a temperature rise +2 °C, (c) °C 20% snow cover. In scenario (b), results showed increased 53% summer (April–September). contrast, projected discharge for scenarios 37% 67%, respectively. efficiently forecasts future water supplies high elevation, data-scarce mountain environments.

Language: Английский

---

Forecasting of SPI and Meteorological Drought Based on the Artificial Neural Network and M5P Model Tree

Land, Journal Year: 2022, Volume and Issue: 11(11), P. 2040 - 2040

Published: Nov. 14, 2022

Climate change has caused droughts to increase in frequency and severity worldwide, which attracted scientists create drought prediction models mitigate the impacts of droughts. One most important challenges addressing is developing accurate predict their discrete characteristics, i.e., occurrence, duration, severity. The current research examined performance several different machine learning models, including Artificial Neural Network (ANN) M5P Tree forecasting widely used measure, Standardized Precipitation Index (SPI), at both time scales (SPI 3, SPI 6). model was developed utilizing rainfall data from two stations India (i.e., Angangaon Dahalewadi) for 2000–2019, wherein first 14 years are employed training, while remaining six validation. subset regression analysis performed on 12 input combinations choose best combination 3 6. sensitivity carried out given find effective parameter forecasting. all ANN (4, 5), (5, 6), (6, 7), assessed through statistical indicators, namely, MAE, RMSE, RAE, RRSE, r. results revealed that (t-1) sensitive parameters with highest values β = 0.916, 1.017, respectively, SPI-3 SPI-6 7 (SPI-1/SPI-3/SPI-4/SPI-5/SPI-8/SPI-9/SPI-11) 4 (SPI-1/SPI-2/SPI-6/SPI-7) based higher R2 Adjusted lowest MSE values. It clear r lesser RMSE as compared 7) models. Therefore, superior other stations.

Language: Английский

---