Robust Hybrid Deep Learning for Potato Price Forecasting in Agricultural Management

With the continuous development of the Internet of Things, artificial intelligence, big data technology, and intelligent agriculture have become hot topics in agricultural science and technology research. Machine learning is one of the core topics in artificial intelligence, and its application has penetrated every aspect of human social life. In modern agricultural intelligent management and decision making, machine learning plays an important role in crop classification, crop disease and insect pest prediction, agricultural product price prediction, and other aspects of management and decision-making processes in agriculture. To detect and recognize the latest research developing features in a quantitative and visual way, and based on machine learning methods in agricultural management, the authors of this paper used CiteSpace bibliometric methods to analyze relevant studies on the development process and hot spots. High-value references, productive authors, country and institution distributions, journal visualizations, research topics, and emerging trends were reviewed and analyzed. According to the keyword visualization and high-value references, machine learning approaches focus on sustainable agriculture, water resources, remote sensing, and machine learning methods. The research mainly focuses on six topics: learning technology, land environment, reference evapotranspiration, decision support systems for river geography, soil management, and winter wheat, while learning technology has been the most popular in recent years.

Adoption of Unmanned Aerial Vehicle (UAV) imagery in agricultural management: A systematic literature review

Improving crop type mapping by integrating LSTM with temporal random masking and pixel-set spatial information

Natural environments face substantial challenges from human activities related to food, feed, and energy production. Unsustainable nutrient management is a key issue, with excess nutrients leaching into the groundwater cycle or escaping intended cropland through other pollution pathways ending up in the atmosphere or in nearby coastal systems. This nutrient loss depletes soil health, contributes to the climate crisis and impacts water quality, especially when combined with intensive farming practices lacking conservation efforts. Innovative mitigation actions, such as the Nature-based Solutions framework, designed to enhance water quality and advance sustainability in agricultural management, require thorough assessment and monitoring to encourage stakeholder participation in these strategies. Conducting research to explore the extent of their effects is thus essential, with a deeper understanding of the nutrient cycle playing a pivotal role in achieving these goals.With the cumulatively increasing availability of remote sensing data sources and advancements in machine learning technologies, automating monitoring and assessment efforts has become a hot and important topic. The challenge is to construct transparent and transferable models capable of working with real-time data to accurately predict crop types, crop status or other desired features. The primary goal of this study is to investigate how an automated multisource data analysis approach, with a focus on remotely sensed data, can support the quantification and mapping of sustainability efforts in agricultural crop management while enhancing the understanding of nutrient flow within large-scale agricultural catchments. Centered on the Hjarbæk Fjord in Denmark, the study also aims to assess the transferability of its models across different sites in Europe. This research is part of a broader project investigating the potential of integrating permanent grasslands into crop rotations as a Nature-based Solution in the catchments surrounding Hjarbæk Fjord. The project aims to develop a decision support tool to guide the planning and optimization of grassland implementation in terms of extend and location. This tool is designed to maximize benefits across various parameters, including the number of stakeholders impacted, economic considerations, crop yield, biodiversity, and other critical factors. The output of the current study, involving the training of a deep learning model to predict cropland trends related to grassland implementation, can in turn be integrated as input for the described decision support tool.This is an explorative study that relies on the availability of accurate ground truth data to train and validate a deep learning model, providing insights into trends associated with the implementation of sustainable management strategies. A key challenge lies in acquiring knowledge of and access to comprehensive datasets that capture relevant parameters, such as actual yield values, quantitative values of nutrients in different stages of the growth season and different nutrient pools within the cropland environment, accurate accounts of management actions and other contributors to the nutrient cycle. Additional challenges involve preprocessing satellite data to establish a robust pipeline for the automated collection of satellite imagery, ensuring a coherent time series. This includes addressing temporal and spatial data gaps through extrapolated estimations to create a consistent dataset.

A new method of farm management that makes use of cutting-edge information technology is known as precision agriculture management. By reducing the waste of water, fertilizers, pesticides, fuel, and other types of agricultural inputs, as well as by optimizing agrarian income and limiting adverse effects on the environment, precision agriculture management seeks to boost agricultural output and quality. Traditional agricultural management methods often lead to inefficient use of resources, higher environmental impacts, and decreased crop yields. Finding efficient, scalable, and accurate ways to track important agricultural variables over large regions is a huge challenge. This paper proposes a new framework called ARS-DLConvNN to handle these issues and enhance the management of agritech and precise agriculture. This framework Integrates High-Resolution Agricultural Remote Sensing (ARS) data and Deep Learning (DL) methods like Convolutional Neural Networks (ConvNN). This approach uses Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) network specifically designed to handle hyperspectral data collected from UAVs and multi-spectral satellite photos. These algorithms can assess crop stress in real-time, provide recommendations on effectively managing water, fertilizer, and insects, and learn to estimate weight and yield. It will be quite easy to see how the proposed method improves agriculture management effectiveness and harvest yields. Improved agricultural management efficiency and crop production were striking when using the suggested deep-learning remote sensing system. On average, crop yields increased by 18% compared to traditional methods, according to field trials in various locations. decrease output. With a 92% success rate in detecting crop illnesses early and an 89% success rate in forecasting water stress, the model allowed for prompt treatments.

Accurate evaluation of evapotranspiration (ET) is crucial for efficient agricultural water management. Data‐driven models exhibit strong predictive ET capabilities, yet significant limitations like naive extrapolation hamper wider generalization. In this perspective, we explore a novel hybrid deep learning (DL) framework to integrate domain knowledge and demonstrate its potential for evaluating ET under the influence of soil salinity. Specifically, we integrated physical constraints from process models (Penman‐Monteith or Shuttleworth‐Wallace) and salinity‐induced stomatal stress mechanisms into the DL algorithm, and evaluated its performance by comparing four diverse scenarios. Results demonstrate that hybrid DL framework offers a promising alternative for ET estimation, achieving comparable accuracy to pure DL during training and validation. Nonetheless, due to the limited available measurements, data‐driven model may not adequately capture plant responses to salt stress, leading to significant prediction biases observed during independent testing. Encouragingly, the hybrid DL model (DL‐SS) integrating Shuttleworth‐Wallace and salinity‐induced stomatal stress mechanisms demonstrated enhanced interpretability, generalizability, and extrapolation capabilities. During testing, DL‐SS consistently showed optimal performance, yielding root mean square error (RMSE) values of 37.4 W m−2 for sunflower and 39.2 W m−2 for maize. Compared to traditional Jarvis‐type approaches (JPM and JSW) and pure DL model during testing, DL‐SS achieved substantial reductions in RMSE values: 51%, 33%, and 43% for sunflower, and 45%, 31%, and 35% for maize, respectively. These findings highlight the importance of integrating prior scientific knowledge into data‐driven models to enhance extrapolation capability of ET modeling, especially in salinized regions where conventional models may struggle.

Remotely sensed soil moisture forecasting through satellite-based sensors to estimate the future state of the underlying soils plays a critical role in planning and managing water resources and sustainable agricultural practices. In this paper, Deep Learning (DL) hybrid models (i.e., CEEMDAN-CNN-GRU) are designed for daily time-step surface soil moisture (SSM) forecasts, employing the gated recurrent unit (GRU), complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), and convolutional neural network (CNN). To establish the objective model’s viability for SSM forecasting at multi-step daily horizons, the hybrid CEEMDAN-CNN-GRU model is tested at 1st, 5th, 7th, 14th, 21st, and 30th day ahead period by assimilating a comprehensive pool of 52 predictor dataset obtained from three distinct data sources. Data comprise satellite-derived Global Land Data Assimilation System (GLDAS) repository a global, high-temporal resolution, unique terrestrial modelling system, and ground-based variables from Scientific Information Landowners (SILO) and synoptic-scale climate indices. The results demonstrate the forecasting capability of the hybrid CEEMDAN-CNN-GRU model with respect to the counterpart comparative models. This is supported by a relatively lower value of the mean absolute percentage and root mean square error. In terms of the statistical score metrics and infographics employed to test the final model’s utility, the proposed CEEMDAN-CNN-GRU models are considerably superior compared to a standalone and other hybrid method tested on independent SSM data developed through feature selection approaches. Thus, the proposed approach can be successfully implemented in hydrology and agriculture management.

Surface soil moisture (SSM) and root-zone soil moisture (RZSM) are key hydrological variables for the agricultural water cycle and vegetation growth. Accurate SSM and RZSM forecasting at sub-seasonal scales would be valuable for agricultural water management and preparations. Currently, weather model-based soil moisture predictions are subject to large uncertainties due to inaccurate initial conditions and empirical parameterization schemes, while the data-driven machine learning methods have limitations in modeling long-term temporal dependences of SSM and RZSM because of the lack of considerations in the soil water process. Thus, here, we innovatively integrate the model-based soil moisture predictions from a sub-seasonal-to-seasonal (S2S) model into a data-driven stacked deep learning model to construct a hybrid SSM and RZSM forecasting framework. The hybrid forecasting model is evaluated over the Yangtze River Basin and parts of Europe from 1- to 46-day lead times and is compared with four baseline methods, including the support vector regression (SVR), random forest (RF), convolutional long short-term memory (ConvLSTM) and the S2S model. The results indicate substantial skill improvements in the hybrid model relative to baseline models over the two study areas spatiotemporally, in terms of the correlation coefficient, unbiased root mean square error (ubRMSE) and RMSE. The hybrid forecasting model benefits from the long-lead predictive skill from S2S and retains the advantages of data-driven soil moisture memory modeling at short-lead scales, which account for the superiority of hybrid forecasting. Overall, the developed hybrid model is promising for improved sub-seasonal SSM and RZSM forecasting over global and local areas.

<b><sc>Abstract.</sc></b> The cocoa pod borer (CPB) (Conopomorpha cramerella) is a very small insect pest native to Asia and Oceania. CPBs cause extensive damage by boring holes into cocoa pod husks and cause premature ripening. Due to its resemblance to other insect pest species, most farm managers fail to recognize it; this makes farm managers unable to avert crop damage. This shows that an automated method for counting CPBs is necessary to allow farm managers to perform integrated pest management (IPM) more effectively. This research proposes a lightweight deep learning algorithm for the on-site counting of CPBs on scanned sticky paper trap images. Sticky paper traps were placed on cocoa plantations to monitor the presence of CPBs. Each sticky paper trap image was obtained using a flatbed scanner to form a dataset called CPB1722. A deep learning model was trained and used to detect the CPBs on sticky paper trap images while bounding box analysis was applied as a lightweight approach to improve overall algorithm performance. The proposed algorithm can detect CPBs on the sticky paper trap images with an F₁-score of 0.89 and a R² of 0.98, relative to the number of manually counted CPBs. The algorithm was tested using an edge device with an average computation time of 24 seconds per image, which was fast enough for the on-site detection of CPBs. The developed algorithm can be used to build a portable imaging device for seamless counting of insects on sticky paper traps.

Artificial intelligence (AI) plays an essential role in agricultural mapping. It reduces costs and time and increases efficiency in agricultural management activities, which improves the food industry. Agricultural mapping is necessary for resource management and requires technologies for farming challenges. The mapping in agricultural AI applications gives efficiency in mapping and its subsequent use in decision-making. This study analyses AI’s current state in agricultural mapping through bibliometric indicators and a literature review to identify methods, agricultural resources, geomatic tools, mapping types, and their applications in agricultural management. The methodology begins with a bibliographic search in Scopus and the Web of Science (WoS). Subsequently, a bibliographic data analysis and literature review establish the scientific contribution, collaboration, AI methods, and trends. The United States (USA), Spain, and Italy are countries that produce and collaborate more in this area of knowledge. Of the studies, 76% use machine learning (ML) and 24% use deep learning (DL) for agricultural mapping applications. Prevailing algorithms such as Random Forest (RF), Artificial Neural Networks (ANNs), and Support Vector Machines (SVMs) correlate mapping activities in agricultural management. In addition, AI contributes to agricultural mapping in activities associated with production, disease detection, crop classification, rural planning, forest dynamics, and irrigation system improvements.

A survey on using deep learning techniques for plant disease diagnosis and recommendations for development of appropriate tools

Continuous observation and management of agriculture are essential to estimate crop yield and crop failure. Remote sensing is cost-effective, as well as being an efficient solution to monitor agriculture on a larger scale. With high-resolution satellite datasets, the monitoring and mapping of agricultural land are easier and more effective. Nowadays, the applicability of deep learning is continuously increasing in numerous scientific domains due to the availability of high-end computing facilities. In this study, deep learning (U-Net) has been implemented in the mapping of different agricultural land use types over a part of Punjab, India, using the Sentinel-2 data. As a comparative analysis, a well-known machine learning random forest (RF) has been tested. To assess the agricultural land, the major winter season crop types, i.e., wheat, berseem, mustard, and other vegetation have been considered. In the experimental outcomes, the U-Net deep learning and RF classifiers achieved 97.8% (kappa value: 0.9691) and 96.2% (Kappa value: 0.9469), respectively. Since little information exists on the vegetation cultivated by smallholders in the region, this study is particularly helpful in the assessment of the mustard (Brassica nigra), and berseem (Trifolium alexandrinum) acreage in the region. Deep learning on remote sensing data allows the object-level detection of the earth’s surface imagery.

It is a big challenge to realize accurate security detection of blast furnace bearing at the same time so as to guarantee the security of equipment. To end this problem, this paper proposed a computer vision technology based on sensor data and hybrid deep learning method for the solution. We use Variational Mode Decomposition (VMD) algorithm which is a new time-frequency analysis method, which can decompose multi-component signals into multiple single-component amplitude-modulated signals at one time to decompose and deal with the sensor data of bearing fault, so as to realize the effective stripping of fault components and original components from sensor data. Using the artificial intelligence mentioned above, the features can be quickly and accurately extracted. By combining the advantages of deep learning, we improve the coupling mechanism and implement a hybrid deep learning-based computer vision method which greatly improves the calculation speed and accuracy of bearing fault diagnosis. It can be fully connected with the feature extraction algorithm VMD, which overcomes the problem that the bearing feature component is easy to be submerged and difficult to extract under the condition of high temperature and strong noise. The results show that the optimal selection of parameters of computer vision technology based on sensor data and hybrid deep learning can be realized through training the sensor data obtained from the experiment. The optimized hybrid deep learning-based computer vision algorithm can achieve 97.4% bearing fault diagnosis hit rate, which is an advanced application of deep learning algorithm in the engineering field.

The project,” CropShield: Price Prediction and Disease Detection for Smart Pesticide Advisory” provides an AIpowered system to support farmers and officials in decision making. It features crop price prediction, offering future market estimates based on historical and regional data, and leaf disease detection using CNNs to identify diseases from images and recommend treatments. Talathi contribute regional data, ensuring accurate predictions. The platform also integrates weather forecasts and government schemes, enhancing its utility. With a user-friendly interface and con tinuous model refinement, this solution aims to boost crop yield, reduce losses, and improve price management in agriculture, ultimately advancing productivity and economic stability

Optimizing agricultural water resource management is crucial for food production, as effective water management can significantly improve irrigation efficiency and crop yields. Currently, precise agricultural water demand forecasting and management have become key research focuses; however, existing methods often fail to capture complex spatial and temporal dependencies. To address this, we propose a novel deep learning framework that combines remote sensing technology with the UNet-ConvLSTM (UCL) model to effectively integrate spatial and temporal features from MODIS and GLDAS datasets. Our model leverages the high-resolution spatial data from UNet and the temporal dependencies captured by ConvLSTM to significantly improve prediction accuracy. Experimental results demonstrate that our UCL model achieves the best R2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$R^2$$\\end{document} compared to existing methods, reaching 0.927 on the MODIS dataset and 0.935 on the GLDAS dataset. This approach highlights the potential of AI and remote sensing technologies in addressing critical challenges in agricultural water management, contributing to more sustainable and efficient food production systems.