Agricultural Fields Research Articles

PEMBERDAYAAN PETERNAK MELALUI PELATIHAN PEMBUATAN PUPUK KOMPOS DESA SOROPIA KECAMATAN SOROPIA KABUPATEN KONAWE

Community service activities integrated with KKN-THEMATIC carried out by the UHO Faculty of Animal Husbandry Lecturer Team took place in Soropia Village, Soropia District, Konawe Regency. Compost is an organic fertilizer produced through the process of decomposition of organic materials, especially animal waste, by active microorganisms such as bacteria and fungi. The purpose of this service is to provide training and technology transfer in making compost. The methods used include delivering materials and demonstrations. In this activity, training in making compost was carried out by utilizing cow dung which is widely available around the Soropia Village settlement, which is then mixed with super decomposer EM4, rice husks, granulated sugar, and water. Through this training, the community is given knowledge about the use of waste from agriculture and livestock to be processed into useful products, especially in the fields of livestock and organic farming which are currently developing in Soropia Village. The farmers showed high enthusiasm during this activity and appreciated the service carried out, with the hope that similar activities can be continued in the future.

Deep learning networks-based tomato disease and pest detection: a first review of research studies using real field datasets.

Recent advances in deep neural networks in terms of convolutional neural networks (CNNs) have enabled researchers to significantly improve the accuracy and speed of object recognition systems and their application to plant disease and pest detection and diagnosis. This paper presents the first comprehensive review and analysis of deep learning approaches for disease and pest detection in tomato plants, using self-collected field-based and benchmarking datasets extracted from real agricultural scenarios. The review shows that only a few studies available in the literature used data from real agricultural fields such as the PlantDoc dataset. The paper also reveals overoptimistic results of the huge number of studies in the literature that used the PlantVillage dataset collected under (controlled) laboratory conditions. This finding is consistent with the characteristics of the dataset, which consists of leaf images with a uniform background. The uniformity of the background images facilitates object detection and classification, resulting in higher performance-metric values for the models. However, such models are not very useful in agricultural practice, and it remains desirable to establish large datasets of plant diseases under real conditions. With some of the self-generated datasets from real agricultural fields reviewed in this paper, high performance values above 90% can be achieved by applying different (improved) CNN architectures such as Faster R-CNN and YOLO.

Estimating Winter Canola Aboveground Biomass from Hyperspectral Images Using Narrowband Spectra-Texture Features and Machine Learning.

Aboveground biomass (AGB) is a critical indicator for monitoring the crop growth status and predicting yields. UAV remote sensing technology offers an efficient and non-destructive method for collecting crop information in small-scale agricultural fields. High-resolution hyperspectral images provide abundant spectral-textural information, but whether they can enhance the accuracy of crop biomass estimations remains subject to further investigation. This study evaluates the predictability of winter canola AGB by integrating the narrowband spectra and texture features from UAV hyperspectral images. Specifically, narrowband spectra and vegetation indices were extracted from the hyperspectral images. The Gray Level Co-occurrence Matrix (GLCM) method was employed to compute texture indices. Correlation analysis and autocorrelation analysis were utilized to determine the final spectral feature scheme, texture feature scheme, and spectral-texture feature scheme. Subsequently, machine learning algorithms were applied to develop estimation models for winter canola biomass. The results indicate: (1) For spectra features, narrow-bands at 450~510 nm, 680~738 nm, 910~940 nm wavelength, as well as vegetation indices containing red-edge narrow-bands, showed outstanding performance with correlation coefficients ranging from 0.49 to 0.65; For texture features, narrow-band texture parameters CON, DIS, ENT, ASM, and vegetation index texture parameter COR demonstrated significant performance, with correlation coefficients between 0.65 and 0.72; (2) The Adaboost model using the spectra-texture feature scheme exhibited the best performance in estimating winter canola biomass (R2 = 0.91; RMSE = 1710.79 kg/ha; NRMSE = 19.88%); (3) The combined use of narrowband spectra and texture feature significantly improved the estimation accuracy of winter canola biomass. Compared to the spectra feature scheme, the model's R2 increased by 11.2%, RMSE decreased by 29%, and NRMSE reduced by 17%. These findings provide a reference for studies on UAV hyperspectral remote sensing monitoring of crop growth status.

Multi-Sensor Soil Probe and Machine Learning Modeling for Predicting Soil Properties.

We present a data-driven, in situ proximal multi-sensor digital soil mapping approach to develop digital twins for multiple agricultural fields. A novel Digital Soil CoreTM (DSC) Probe was engineered that contains seven sensors, each of a distinct modality, including sleeve friction, tip force, dielectric permittivity, electrical resistivity, soil imagery, acoustics, and visible and near-infrared spectroscopy. The DSC System integrates the DSC Probe, DSC software (v2023.10), and deployment equipment components to sense soil characteristics at a high vertical spatial resolution (mm scale) along in situ soil profiles up to a depth of 120 cm in about 60 s. The DSC Probe in situ proximal data are harmonized into a data cube providing vertical high-density knowledge associated with physical-chemical-biological soil conditions. In contrast, conventional ex situ soil samples derived from soil cores, soil pits, or surface samples analyzed using laboratory and other methods are bound by a substantially coarser spatial resolution and multiple compounding errors. Our objective was to investigate the effects of the mismatched scale between high-resolution in situ proximal sensor data and coarser-resolution ex situ soil laboratory measurements to develop soil prediction models. Our study was conducted in central California soil in almond orchards. We collected DSC sensor data and spatially co-located soil cores that were sliced into narrow layers for laboratory-based soil measurements. Partial Least Squares Regression (PLSR) cross-validation was used to compare the results of testing four data integration methods. Method A reduced the high-resolution sensor data to discrete values paired with layer-based soil laboratory measurements. Method B used stochastic distributions of sensor data paired with layer-based soil laboratory measurements. Method C allocated the same soil analytical data to each one of the high-resolution multi-sensor data within a soil layer. Method D linked the high-density multi-sensor soil data directly to crop responses (crop performance and behavior metrics), bypassing costly laboratory soil analysis. Overall, the soil models derived from Method C outperformed Methods A and B. Soil predictions derived using Method D were the most cost-effective for directly assessing soil-crop relationships, making this method well suited for industrial-scale precision agriculture applications.

Smart integration in agriculture: an Arduino-driven rice grain dryer for optimal post-harvest management

In the field of agriculture, the post-harvest stage often presents a multitude of obstacles that have the potential to undermine the quality and longevity of the crop. One of the primary obstacles in rice production is the effective drying of rice grains, which is crucial for preventing problems such as mold growth, discoloration, and unintentional germination. To tackle this issue, we propose an innovative solution: a rice grain drying system driven by an Arduino microcontroller. This device effectively combines technology accuracy with agricultural needs. The system is equipped with a very precise temperature sensor to maintain a uniform drying environment, fluctuating within the range of 50–60 °C. The use of a nichrome heater, selected for its consistent and dependable heat production, is complemented by an air blower to ensure the even distribution of heat throughout the drying chamber. One notable characteristic of this innovation is its significant drying capacity of 50 kg in each operation. The effectiveness of the product was verified by a rigorous testing process. A batch of glutinous rice weighing 25 kg, initially containing 25.5% moisture, was successfully reduced to 13.5% moisture content within a time frame of 125 min. In a similar vein, a batch of grains weighing 40 kg, which had just undergone threshing, had a moisture content of 22% initially. Remarkably, this batch achieved the desirable moisture level of 14% within a mere two-hour timeframe. This innovative technology not only provides a resolution to the difficulties faced in post-harvest processes but also signifies a fundamental change in thinking, introducing a fresh era where intelligent technology seamlessly integrates with traditional agricultural methods.

Isolation and Characterization of Arsenic-Tolerable Bacteria from Groundwater and Their Implementation on Rice Seedling's Shoot and Root Enhancement.

Arsenic exerts detrimental impacts on primary metabolism in plants, leading to reduced crop yield. Some arsenic-resistant plant growth-promoting bacteria (PGPB) help plants by providing some plant hormones to sustain their growth and development under arsenic stress. Here, seven different species of Bacillus were isolated from arsenic-contaminated groundwater of West Bengal, India. Those species were capable of growing in the presence of > 3.12g/L arsenate (AsV) and > 0.65g/L arsenite (AsIII) salts and also resist different heavy metals like Cu2+, Fe2+, Co2+, Zn2+, Pb2+, etc. They were susceptible to multiple groups of antibiotics like beta-lactam, aminoglycosides, etc. All species were capable of detoxifying arsenite and influenced rice seedlings' growth in the presence of arsenic salts by their capabilities likenitrogen-fixing ability, phosphate solubilization, indole 3-acetic acid (IAA), gibberellic acid (GA), proline production, etc. Most species helped enhance root and shoot lengths under arsenic stress. These primary findings suggest that those Bacillus spp. could be used as potential bio-fertilizers in arsenic-contaminated agricultural fields.

The housing environment and its effect on Hispanic children with asthma

IntroductionSubstandard housing in Hispanic minority communities has been associated with detrimental effects on children with asthma that is affected by diverse indoor triggers. Identifying those triggers and offering adequate medical treatment and ways to prevent exposure to triggers will help reduce the number of asthma attacks among children who reside in Hispanic minority communities.ObjectiveOur study aimed to identify the influence of substandard housing quality and its effect on asthma symptoms. Our objectives included assessing the impact of building, flooring types, heating sources, window shading, parental smoking behaviors, moisture and mold in the home, ventilation, and owning a pet on asthma symptom occurrence.Materials and methodsOur study was based on a cross-sectional analytical study involving 353 individuals, their parents, and their homes. Logistic regression modeling was conducted to explain the dependence of each asthma symptom on the investigated independent variables, with child age and sex being investigated as covariables. A post hoc test was employed to ensure goodness of fit, and a pseudo-R square was calculated for each regression model. The Hosmer-Lemenshow test was performed post hoc to ensure goodness of fit, and models were preserved if p > 0.05.ResultsThe overall prevalence of asthma-related symptoms in our study sample was 19.37%, and sleeping quality was affected due to asthma in 23% of participating children. We also found that children living with asthma near agricultural fields were more prone to asthma-related symptoms, as well as those children living in homes with excessive moisture and those who kept stuffed toys or pets inside their rooms.ConclusionsOur study highlights how a variety of environmental factors, such as living near an agricultural area and having curtains, increases shortness of breath. Having leaks and excessive moisture in houses increases shortness of breath, wheezing, coughing, and rates of colds. Another important exposure is having a pet inside the house which increases chest tightness. A comprehensive understanding of these factors and how they impacted targeted participants, especially those who are Hispanic minorities and living in substandard housing, is essential for informing the design of effective interventions and improving respiratory health in the home environment.

Design and Use of a Stratum-Based Yield Predictions to Address Challenges Associated with Spatial Heterogeneity and Sample Clustering in Agricultural Fields Using Remote Sensing Data

Machine learning (ML) models trained with remote sensing data have the potential to improve cereal yield estimation across various geographic scales. However, the complexity and heterogeneity of agricultural landscapes present significant challenges to the robustness of ML-based field-level yield estimation over large areas. In our study, we propose decomposing the landscape complexity into homogeneous zones using existing landform, agroecological, and climate classification datasets, and subsequently applying stratum-based ML to estimate cereal yield. This approach was tested in a heterogeneous region in northern Morocco, where wheat is the dominant crop. We compared the results of the stratum-based ML with those applied to the entire study area. Sentinel-1 and Sentinel-2 satellite imagery were used as input variables to train three ML models: Random Forest, Extreme Gradient Boosting (XGBoost), and Multiple Linear Regression. The results showed that the XGBoost model outperformed the other assessed models. Furthermore, the stratum-based ML approach significantly improved the yield estimation accuracy, particularly when using landform classifications as homogeneous strata. For example, the accuracy of XGBoost model improved from R2 = 0.58 and RMSE = 840 kg ha−1 when the ML models were trained on data from the entire study area to R2 = 0.72 and RMSE = 809 kg ha−1 when trained in the plain area. These findings highlight that developing stratum-based ML models using landform classification as strata leads to more accurate predictions by allowing the models to better capture local environmental conditions and agricultural practices that affect crop growth.

An Improved Multi-Scale Feature Extraction Network for Rice Disease and Pest Recognition

In the process of rice production, rice pests are one of the main factors that cause rice yield reduction. To implement prevention and control measures, it is necessary to accurately identify the types of rice pests and diseases. However, the application of image recognition technologies focused on the agricultural field, especially in the field of rice disease and pest identification, is relatively limited. Existing research on rice diseases and pests has problems such as single data types, low data volume, and low recognition accuracy. Therefore, we constructed the rice pest and disease dataset (RPDD), which was expanded through data enhancement methods. Then, based on the ResNet structure and the convolutional attention mechanism module, we proposed a Lightweight Multi-scale Feature Extraction Network (LMN) to extract multi-scale features at a finer granularity. The proposed LMN model achieved an average classification accuracy of 95.38% and an F1-Score of 94.5% on the RPDD. The parameter size of the model is 1.4 M, and the FLOPs is 1.65 G. The results suggest that the LMN model performs rice disease and pest classification tasks more effectively than the baseline ResNet model by significantly reducing the model size and improving accuracy.

Assessment of microplastic and heavy metal pollution in agricultural soils of Ernakulam District, Kerala, India.

Microplastics (MPs) and heavy metal pollution pose significant environmental threat, potentially leading to agroecosystem toxicity and jeopardizing food security. Therefore, this study aims to evaluate the abundance and risk assessment of these pollutants in five different farmlands of Ernakulam district, India. Results showed that MPs content in agricultural fields near commercialized areas such as Kakkanad Nedungapuzha, Nedumbassery, and Kadamakuddy was dominant compared to Nechoor, a rural area. The average microplastic abundance was found to be 45.6 ± 26.4 items kg⁻1 dw. Polypropylene (PP) and polyethylene (PE) were the dominant polymers found in the soil samples, constituting 45% and 25% of the microplastic content, respectively. The pollution load index of MPs indicates that the sampling sites are considered to be polluted as PLI > 1 with hazard level I. The heavy metal pollution status follows the order: Cu (80.3 to 724mg/kg) > Zn (77 to 543.5mg/kg) > Cr (171.65 to 334.65mg/kg) > As (10.25 to 79.5mg/kg) > Pb (2.05 to 30.3mg/kg) > Cd (0.3 to 14.35mg/kg). Calculated pollution load index (PLI) geo-accumulation index (Igeo), ecological risk assessment values indicate that commercialized regions exhibit high levels of trace metals, namely Cu, Zn, As, Cd, and Cr, posing a significant concern for the agricultural ecosystem. Our results indicate heightened microplastics and heavy metals prevalence in farmlands adjacent to commercial zones, necessitating immediate preventive action to mitigate increasing concentrations.

Blockchain adoption factors in agricultural supply chains: A PLS-SEM study

The purpose of this study is to explore factors influencing the blockchain adoption in agricultural supply chains, to make a particular focus on how security and privacy considerations, policy support, and management support impact the blockchain adoption intention. it further investigates perceived usefulness as a mediating variable that potentially amplifies the effects of these factors on blockchain adoption intention, and sets perceived cost as a moderating variable to test its influence on the strength and direction of the relationship between perceived usefulness and adoption intention. through embedding the cost-benefit theory into the integrated tam-toe framework and utilizing the partial least squares structural equation modeling (PLS-SEM) method, this study identifies the pivotal factors that drive or impede blockchain adoption in the agricultural supply chains, which fills the gap of the relatively insufficient research on the blockchain adoption in agriculture field. the results further provide empirical evidence and strategic insights that can guide practical implementations, to equip stakeholders or practitioners with the necessary knowledge to navigate the complexities of integrating cutting-edge technologies into traditional agricultural operations, thereby promoting more efficient, transparent, and resilient agricultural supply chains.

Potassium limits productivity in intensive cereal cropping systems in Southeast Asia.

Potassium (K) has received little attention as a potential yield-limiting factor in cropping systems. Here we investigated the K status in intensive cereal cropping systems in Indonesia, which are representative of many other Southeast Asian countries. Our analysis included nutrient input-output balance, leaf nutrient status, long- and short-term fertilizer trials, and farmer surveys. We revealed that soil K levels alone are insufficient to meet plant requirements, and current fertilizer applications are inadequate to prevent K deficiencies and large negative annual soil K balances in farmer fields (average -62 kgK ha-1). On-farm fertilizer trials indicated that nearly 80% of rice crops and 70% of maize crops achieved higher yields with the application of K fertilizer. Addressing K limitations will require an enhanced capacity to predict crop responses to K fertilizer, together with long-term, flexible fertilizer and crop residue management strategies. Furthermore, similar K limitations have probably emerged in other regions globally due to intensive cropping with insufficient K replenishment, which must be addressed to close yield gaps on existing farmland.

Hydrochemical Characterisation of the Basement Aquifer with a Focus on the Origin of Nitrate in the Highly Urbanised Niamey Region, SW of Niger

This study investigated the basement aquifer beneath the urbanised city of Niamey and the agricultural fields of Kollo, SW of Niger. During the observation period spanning from 2021 to 2023, groundwater and surface water samples were collected for analysis of major ions and the stable isotopes oxygen-18 and deuterium (δ18O and δ2H) of water. To trace the origin of high nitrate concentrations (NO3−) found in several observation and drinking water wells in both areas, δ15N and δ18O isotope values of NO3− were analysed in groundwater and eluted soil samples. The observed hydrochemical patterns mainly reflect the heterogeneity of the weathered fringe of the basement aquifer. Decreasing concentrations of NO3− and δ18O and δ2H values were observed in relation to the distance of the Niger River and increasing thickness of the clay layer on the surface. The wells close to the river in Niamey show a dilution effect during the flood season, and the NO3− concentrations displayed a continuous increasing trend. The δ15N-NO3 and δ18O-NO3 values confirmed that septic tank water is spreading in the region of Niamey and that manure originating from livestock in Kollo is the main source of NO3−. The patterns of δ15N in the soil samples coincide with those of cattle’s manure spread in both areas. The shallow wells show significantly higher values of electric conductivity and NO3− concentrations compared to the deeper wells, which clearly indicates the influence of shallow septic tanks on water quality.

Agricultural Valorization of Treated Effluent from Korhogo Wastewater Treatment Plant

Aims: The main objective is to evaluate the effluents of Korhogo’s wastewater treatment plant for irrigation and fertilization in agriculture. Study Design: Characterization of outlet treated wastewaters has been firstly done. Then their suitability for irrigation has been determined and finally their fertilizing power has been assessed. Place and Duration of Study: Department of geosciences, Korhogo’s wastewater treatment plant and laboratory. Sampling and analyses were carried out in January 2023, before agricultural field implementation started in February 2023. Methodology: Firstly, Physico-chemical and microbiological analyses, allowed characterization of outlet treated wastewaters. A determination of their suitability for irrigation was then done through Sodium Adsorption Ratio calculation and water classification with Diagram of Richards. An assessment of their fertilizing capacity was finally carried out using these waters for irrigating and fertilizing crop plots (3x2 m), and consisting of determining firstly their biodegradability index, then their nutrient contents (Nitrogen, Phosphorus and potassium), and the influence of their use on crop development and yield. Results: The results show that several parameters meet standards for discharge of treated wastewater except total nitrogen, suspended matters and total phosphorus (67.67 mg/l, 271.33 mg/l and 18.30 mg/l). Sodium Adsorption Ratio (SAR) value=3.78 and C3S1 class indicate that these treated effluents can be used for irrigation without salinization risk. As fertilizer, treated effluents brought to crops respectively 603 kg/ha of nitrogen, 162 kg/ha of phosphorus and 99 kg/ha of potassium for maize; and 368.5 kg/ha of nitrogen, 99 kg/ha of phosphorus and 60.5 kg/ha of potassium respectively for Okra. A yield of 12.2 kg of fresh maize and 3.15 kg of fresh okra was obtained for the plot irrigated by treated effluents, compared with 4.80 kg of fresh maize and 0.36 kg of fresh okra for the plot irrigated by well water. Conclusion: This study has shown that outlet treated wastewater from Korhogo’s wastewater treatment plant is suitable for crops irrigation, contains sufficient nutrients enhancing crops development and yield.

Public-private partnerships in the agricultural sector

The levels of management in the field of agriculture were systematized, a comparative ana­lysis of the international experience of management in the agrarian sphere was carried out (on the example of Germany and Italy). The current trends in digitalization of agricultural management in the world, as well as tasks and challenges of digitalization in Ukraine were described. The hypothesis of the research is that the implementation of digital solutions in the Ukrainian agricultural sector will increase the efficiency of management, especially in the condi­tions of military challenges. Methodo­logically, the research is based on the analysis of policies, strategic plans, scientific works and other sources related to management in the field of agriculture, taking into account the impact of digital techno­logies. The centralized manage­ment model in the development of the agricul­tural sector still dominates in European coun­tries. However, the EU current policy in the field of regional development provides for the transfer of as many management functions as possible to the region­nal level. This approach is also characteristic of the national model of agricultural development policy. The most important and effective tool for the modernization of any of the known management systems in the modern world is their digitali­zation. Digitization of management in the agricu­ltural industry at all levels, from the state to small agribusiness entities, is no exception. The further digital transformation of manage­ment in national agriculture represents a higher level of digital integration, which affects the most complex organi­zational changes in the structures of state admi­nistration and agribusi­ness. The results of the implementation of these tasks can significantly affect the growth of profits in the agricultural business and the competitiveness of products, which will allow the agricultural industry to reach modern global technological boundaries

Optimization, experimental and theoretical investigation of modified guar-gum and poly(acrylic acid) based smart hydrogel for controlled release applications in the agricultural field

A series of smart polymer hydrogels with good uptake and controlled release characteristics were synthesized free-radically by growing poly(acrylic acid) on allylated guar gum using ammonium persulfate as an initiator and trimethylolpropane triacrylate as the crosslinker. The synthesized hydrogels were evaluated using FT-IR and NMR for their structure, thermo-gravimetric analysis (TG) to measure thermal stability, and scanning electron microscopy (SEM) to analyze surface morphology and swellability by gravimetric method under different pH and temperature conditions. The soil moisture content and gel fractions were measured using gravimetric methods. Additionally, a theoretical investigation was also performed to assess the urea-holding capacity of the synthesized hydrogel matrix using density functional theory (DFT) calculations. The uptake study revealed that the hydrogel with optimized monomer feed composition absorbed 851 g/g of water, 78 % and 67 % of urea from 0.5 % and 1 % urea solutions respectively under ambient conditions. The release study showed that 56 and 47 % of absorbed urea were released over 15 days from the urea-loaded hydrogels (0.5 % and 1 % urea solutions respectively). The uptake mechanism obeyed the non-Fickian and first-order kinetics models. The experimental evaluation and DFT calculations conveyed that the synthesized hydrogel could act as a candidature matrix for the controlled release of water and fertilizer in the agricultural field.

Crop Water Stress Index and Yield Relationships for Winter Wheat (Triticum aestivum) Crops Grown Under Different Drip and Flood Irrigated Treatments

ABSTRACTThe Crop Water Stress Index (CWSI) is a widely used method for quantifying crop water status and predicting yield. However, its evaluation across different irrigation methods and its stage‐specific response to crop yield is rarely evaluated. In this study, controlled field experiments were conducted on winter wheat using drip irrigation (DI) and flood irrigation (FI) during the 2021–2022 and 2022–2023 seasons in western Uttar Pradesh, India. The irrigation treatments included 50% MAD (maximum allowable depletion) (DI), 55% MAD (DI), 60% MAD (DI), 50% MAD (FI), local farmer's field replication (FI), rain‐fed, and well‐watered treatment (DI). The derived mean CWSI values for the irrigation treatments ranged from 0.03 to 0.66 in season 1 and 0.06 to 0.57 in season 2 across treatments. The seasonal mean CWSI for 50% MAD (DI) was 0.12 (season 1) and 0.11 (season 2), while 50% MAD (FI) yielded higher mean CWSI values of 0.29 (season 1) and 0.22 (season 2). The 50% MAD (DI) treatment produced the highest grain yield and water use efficiency in both seasons. A comprehensive analysis of stage‐specific CWSI values and grain yields revealed that grain yield was more sensitive to post‐heading CWSI as compared to pre‐heading CWSI values. Among the growth stages, CWSI values during the flowering stage were the most critical for predicting wheat yield. The study recommends that the CWSI values in the flowering and post‐heading stages are more relevant in predicting wheat yield accurately as compared to the pre‐heading and seasonal mean CWSI.

Flooded area detection and mapping from Sentinel-1 imagery. Complementary approaches and comparative performance evaluation

ABSTRACT The current study assesses the performance of several machine learning (ML) and deep learning (DL) models for detecting and mapping floods using Sentinel-1 SAR imagery. Three distinct approaches were used: pixel classification, object-based image analysis and object instance segmentation. The ML models are Random Forest, and Support Vector Machine and the DL models are U-NET, DeepLabV3 and Mask RCNN. Five different case studies were selected to test the models’ scalability. These areas are in Romania (Prut River, at the border between Romania, the Republic of Moldova and Ukraine, Timiș River, and Râul Negru River), the United States of America (Missouri River) and Australia (Broughton Creek). For each flood, five Sentinel-1 images were used, four collected before the flood and one collected after the flood. The intensity images were stacked and scaled in the range of the intensity thresholds associated with water and non-water so that all the case studies have the same margins for intensity. Samples of water, vegetation, agricultural fields, and bare soil were collected only from the Prut River case study and used in the training process. Out of all models, the U-Net model returned the highest accuracy with a value for Intersect over Union of 0.763 for a tile size of 128x128 pixels.

Synthetic meets authentic: Leveraging LLM generated datasets for YOLO11 and YOLOv10-based apple detection through machine vision sensors

Training machine learning (ML) models for artificial intelligence (AI) and computer vision-based object detection process typically requires large, labeled datasets, a process often burdened by significant human effort and high costs associated with imaging systems and image acquisition. This research aimed to simplify image data collection for object detection in orchards by avoiding traditional fieldwork with different imaging sensors. Utilizing OpenAI's DALLE, a large language model (LLM) for realistic image generation, we generated and annotated a cost-effective dataset. This dataset, exclusively generated by LLM, was then utilized to train two state-of-the-art deep learning models: YOLOV10 and YOLO11. The YOLO11 model for apple detection was trained with its five configurations (YOLO11n, YOLO11 s, YOLO11 m, YOLO11l and YOLO11x), and YOLOv10 model with its six configurations (YOLOv10n, YOLOv10 s, YOLOv10 m, YOLOv10b, YOLOv10l and YOLOv10x), which was then tested with real-world (outdoor orchard) images captured by a digital (Nikon D5100) camera and a consumer RGB-D camera (Microsoft Azure Kinect). YOLO11 outperformed YOLOv10 as YOLO11x and YOLO11n exhibited superior precision of 0.917 and 0.916, respectively. Furthermore, YOLO11l demonstrated the highest recall among its counterparts, achieving a recall of 0.889. Likewise, the YOLO11n variant excelled in terms of mean average precision (mAP@50), achieving the highest value of 0.958. Validation tests against actual images collected through a digital camera (Nikon D5100) over Scilate apple variety in a commercial orchard environment showed a highest precision of 0.874 for YOLO11 s, recall of 0.877 for YOLO11l and mAP@50 of 0.91 for YOLO11x. Additionally, validation test against actual images collected through a Microsoft Azure camera over the same orchard showed a highest precision, recall and mAP@50 respectively of 0.924, 0.781 and 0.855 with YOLO11x. All variants of YOLO11 surprisingly demonstrated a pre-processing time of just 0.2 milliseconds (ms), which was faster than any variant of YOLOv10. The fastest inference time for the YOLO11n model using the training dataset generated by the language model was 3.2 ms, while YOLOv10n, fastest among YOLOv10 variants, had a longer inference time of 5.5 ms. Likewise, the fastest inference time for the sensor-based images was 7.1 ms (for Nikon D5100 camera images) and 4.7 ms (for Azure images) with YOLO11n. This study presents a pathway for generating large image datasets using LLM in challenging agricultural fields with minimal or no labor-intensive efforts in field data-collection, which could accelerate the development and deployment of computer vision and robotic technologies in orchard environments.

A multispectral camera in the VIS–NIR equipped with thermal imaging and environmental sensors for non invasive analysis in precision agriculture

Multispectral imaging (MSI) is a technique used to inspect materials properties in different domains, ranging from industrial to medical and cultural heritage and, recently, precision agriculture. Even though several MSI solutions are already commercially available, the research community is working to optimize multispectral cameras in terms of performance and cost. Systems for the agricultural field are usually very compact, combined with drones for large areas acquisition. In this work, we detail the implementation of an innovative, modular and low-cost solution of a multispectral camera based on three core camera systems in the optical (VIS–NIR) and thermal (LWIR) range. Multispectral imaging is performed with a rotating wheel of interchangeable band-pass filters. The system is also equipped with a set of environmental sensors to acquire CO2 concentration values, light intensity, temperature, and relative humidity of the surrounding environment. The technology and the measurement protocol were experimentally validated in laboratory and in open field. Advantages with respect to the available MSI cameras mounted on UAV is the integrated imaging in both the reflectance and the thermal emissive band in a close-up imaging setup and the use of environmental sensors. From the multispectral stack the spectral signature of the plants can be obtained and various vegetation indices (e.g., NDVI, NDRE) can be calculated for investigating the health status of the plant, while thermography provide additional monitoring. Close-up multispectral imaging is expected to tackle the new challenges of precision agriculture by enabling the acquisition of high-quality dataset on single plants.