AgriEngineering, Volume 7, Issue 7 (July 2025) – 27 articles

The adoption of AI vision models in specialized industries is often hindered by the substantial requirement for extensive, manually annotated image datasets. Even when employing transfer learning, robust model development typically necessitates tens of thousands of such images, a process that is time-consuming, costly, and demands consistent expert annotation. This technical note introduces a semi-automated method to significantly reduce this annotation burden. The proposed approach utilizes two general-purpose vision-transformer-to-caption (GP-ViTC) models to generate descriptive text from images. These captions are then processed by a custom-developed semantic classifier (SC), which requires only minimal training to predict the correct image class. This GP-ViTC + SC system demonstrated exemplary classification rates in test cases and can subsequently be used to automatically annotate large image datasets. While the inference speed of the GP-ViTC models is not suited for real-time applications (approximately 10 s per image), this method substantially lessens the labor and expertise required for dataset creation, thereby facilitating the development of new, high-speed, custom AI vision models for niche applications. This work details the approach and its successful application, offering a cost-effective pathway for generating tailored image training sets. Full article

Background: Understanding canopy light interception is essential for optimizing forage production and improving the efficiency of grazing systems. Accurate quantification of photosynthetically active radiation (PAR) intercepted by the canopy allows for better estimation of crop coefficients and growth dynamics. This study aimed to assess the forage mass and nutritional value of Guinea grass pastures managed under two grazing frequencies, defined by 90% and 95% light interception (LI) measured using AccuPar equipment, and two post-grazing stubble heights (30 and 50 cm). Evaluations were conducted during both the rainy season and a dry year to capture seasonal variability in pasture performance. Methods: The experimental design was of completely randomized blocks with four replications. Results: The treatment whit 90% LI resulted in higher values of crude protein and digestible. However, 95% LI resulted in higher values of neutral detergent insoluble nitrogen and acid detergent insoluble nitrogen values in grass pastures Guinea. The highest value of forage mass in Guinea grass was reported with 95% LI in association with a post-grazing height of 30 cm. Conclusions: Management of light interception at 90% provided a reduced amount of forage with better nutritional value. Pasture management considering the light interception technology with the AccuPar equipment was efficient as a pattern for interrupting pasture regrowth in the vegetative phase. Full article

Strawberry post-harvest losses are estimated at 50%, due to improper handling and harvest timing, necessitating the use of non-invasive methods. This study develops a non-invasive in situ bioelectrical spectroscopy for strawberry peduncles. Based on traditional assessments and invasive metrics, 100 physiologically ripe (PR) and 100 commercially mature (CM) strawberries were distinguished. Spectra from their peduncles were measured from 1 kHz to 1 MHz, collecting four parameters (magnitude (Z(f)), phase angle (θ(f)), resistance (R(f)), and reactance (X(f))), resulting in 80,000 raw data points. Through systematic spectral preprocessing, Bode and Cole–Cole plots revealed a distinction between PR and CM strawberries. Frequency selection identified seven key frequencies (1, 5, 50, 75, 100, 250, 500 kHz) for deriving 37 engineered features from spectral, extrema, and derivative parameters. Feature selection reduced these to 6 parameters: phase angle at 50 kHz (θ (50 kHz)); relaxation time (τ); impedance ratio (|Z₁ₖ/Z₂₅₀ₖ|); dispersion coefficient (α); membrane capacitance (Cm); and intracellular resistivity (ρi). Four algorithms (TabPFN, CatBoost, GPC, EBM) were evaluated with Monte Carlo cross-validation with five iterations, ensuring robust evaluation. CatBoost achieved the highest accuracy at 93.3% ± 2.4%. Invasive reference metrics showed strong correlations with bioelectrical parameters (r = 0.74 for firmness, r = −0.71 for soluble solids). These results demonstrate a solution for precise harvest classification, reducing post-harvest losses without compromising marketability. Full article

Greenhouses are crucial for maintaining an ideal temperature and humidity level for plant growth; however, attaining ideal levels remains a challenge. Energy-efficient and sustainable alternatives are needed because traditional temperature/humidity control practices and vapor compression air conditioning systems depend on climate conditions and harmful refrigerants. Advanced alternative technologies like evaporative cooling and desiccant dehumidification have emerged that maintain the ideal greenhouse temperature and humidity while using the least amount of energy. This study reviews direct evaporative cooling, indirect evaporative cooling, and Maisotsenko-cycle evaporative cooling (MEC) systems and solid and liquid desiccant dehumidification systems. In addition, integrated desiccant and evaporative cooling systems and hybrid systems are reviewed in this study. The results show that the MEC system effectively reduces the ambient temperature up to the ideal range while maintaining the humidity ratio, and both dehumidification systems effectively reduce the humidity level and improve evaporative cooling efficiency. The integrated systems and hybrid systems have the ability to increase energy efficiency and controlled climatic stability in greenhouses. Regular maintenance, initial system cost, economic feasibility, and system scalability are significant challenges to implement these advanced temperature and humidity control systems for greenhouses. These findings will assist agricultural practitioners, engineers, and researchers in seeking alternate efficient cooling methods for greenhouse applications. Future research directions are suggested to manufacture high-efficiency, low-energy consumption, and efficient greenhouse temperature control systems while considering the present challenges. Full article

Multispectral images obtained from Unmanned Aerial Vehicles (UAVs) have become strategic tools in precision agriculture, particularly for analyzing spatial variability in soil attributes. This study aimed to evaluate the spatial distribution of soil electrical (EC) and total organic carbon (TOC) in irrigated forage cactus areas in the Brazilian semiarid region, using field measurements and UAV-based multispectral imagery. The study was conducted in a communal agricultural settlement located in the Mimoso Alluvial Valley (MAV), where EC and TOC were measured at 96 points, and seven biophysical indices were derived from UAV multispectral imagery. Geostatistical models, including cokriging with spectral indices (NDVI, EVI, GDVI, SAVI, and NDSI), were applied to map soil attributes at different spatial scales. Cokriging improved the spatial prediction of EC and TOC by reducing uncertainty and increasing mapping accuracy. The standard deviation of EC decreased from 1.39 (kriging) to 0.67 (cokriging with EVI), and for TOC from 15.55 to 8.78 (cokriging with NDVI and NDSI), reflecting a 43.5% reduction in uncertainty. The indices, EVI, NDVI, and NDSI, showed strong potential in representing and enhancing the spatial variability in soil attributes. NDVI and NDSI were particularly effective at finer grid resolutions, supporting more efficient irrigation strategies and sustainable agricultural practices. Full article

Accurate estimation of fruit quantity is essential for efficient harvest management, storage, transportation, and marketing in the agricultural industry. To address the limited generalizability of single-species models, this study presents a comprehensive deep learning-based framework for multi-species fruit-load estimation, leveraging the MetaFruit dataset, which contains images of five fruit species collected under diverse orchard conditions. Four representative object detection and regression models—YOLOv8, RT-DETR, Faster R-CNN, and a U-Net-based heatmap regression model—were trained and compared as part of the proposed multi-species learning strategy. The models were evaluated on both the internal MetaFruit dataset and two external datasets, NIHS-JBNU and Peach, to assess their generalization performance. Among them, YOLOv8 and the RGBH heatmap regression model achieved F1-scores of 0.7124 and 0.7015, respectively, on the NIHS-JBNU dataset. These results indicate that a deep learning-based multi-species training strategy can significantly enhance the generalizability of fruit-load estimation across diverse field conditions. Full article

The use of artificial intelligence tools in agriculture is growing. In particular, the use of ensemble models. However, there are still few reviews on the use of these models in the study of the impacts of climate change on agriculture. Therefore, the aim of this article is to review the use of such models and perform three key tasks: (1) identify topics in which ensemble models are used, (2) determine the most widely applied model and the predominant crops and regions, and (3) explore future applications and challenges. As a result, it was noted that the first studies, dating back to 2011, applied ensemble models to model invasive species. Since then, research has focused on changes in temperature and precipitation, with at least one study published every year. The most cited studies have dealt with land use classification, emphasizing its relevance to climate change studies. Notably, studies on carbon storage in soil and its capacity to remove CO₂ from the atmosphere have become increasingly relevant. This analysis highlights the growing importance of ensemble models in climate-related agricultural research, outlining trends and key areas for future exploration. Full article

The adsorption performance of maize seeds in air-suction seed metering devices directly affects the operational quality of maize seeders. The suction holes on the seed metering disc play a crucial role in determining the device’s ability to adsorb maize seeds and serve as a key design parameter for air-suction seed metering systems. Existing research has primarily focused on seed posture control and suction force models for standard particles, while experimental studies on the actual adsorption performance of maize seeds remain scarce. To further investigate the adsorption characteristics of maize seeds under different suction hole geometries, this study employed a self-developed adsorption force measurement platform to conduct experiments on maize seeds in various adsorption postures. The resulting force–displacement curves reveal the variation of adsorption force as seeds detach from the suction holes. To assess the applicability of conventional suction force calculation models, computational fluid dynamics (CFD) simulations were performed to analyze the adsorption mechanism of standard particles. The simulation results indicate significant limitations in commonly used suction force estimation methods. For instance, in experiments evaluating the effect of equivalent adsorption area, the relative error between the suction force estimated by the traditional pressure-based method for triangular holes and the actual measured force reached 40.82%. Similarly, the relative error between the force estimated by the airflow drag method for square suction holes and the actual measured force under the same conditions was 17.14%. Therefore, when evaluating actual seed adsorption, it is essential to comprehensively consider factors such as suction hole geometry, blocked suction area, seed shape, vacuum pressure, and the overlap depth between the seed boundary and the suction hole, all of which significantly influence the adsorption effect. Full article

Emitter clogging in drip irrigation systems is a recurring issue, affecting water application uniformity and system lifespan. This study investigated the anti-clogging performance of emitters and the accumulation patterns of solid particles in dripper labyrinths with varied geometric configurations, combining laboratory experimentation and computational fluid dynamics simulations. Fifteen labyrinth models were tested, divided into two groups: (Model A) emitters with well-defined vortexes and (Model B) emitters with uniform flow. The tests were conducted with solid particle concentrations of 125 and 500 mg L⁻¹ over 200 h of operation. The results showed that none of the emitters became clogged, even under severe particle concentration conditions. However, distinct deposition patterns were observed. Emitters with vortex formation accumulated particles in low-velocity zones, especially in the first baffles of the labyrinth. In contrast, emitters with uniform flow minimized sediment buildup, maintaining high velocities throughout the channel section. Simulations confirmed that the relationship between labyrinth geometry and flow velocity directly influences particle deposition. Dripper design strategies aimed at reducing low-velocity zones in the channel could help mitigate clogging risks. The findings of this study provide valuable guidelines for developing more clogging-resistant emitters, contributing to the improvement of drip irrigation systems. Full article

Knowledge of global radiation (Hg) is essential for regional economic development and can help guide public policies related to agricultural and energy potential. However, its availability in several Brazilian regions is still limited. This work evaluates the predictive capacity of two machine learning (ML) techniques, such as multi-layer perceptrons (MLPs) and support vector machines (SVMs), in the estimation of Hg in 20 meteorological stations with 40 different input combinations involving insolation, air temperature, air relative humidity, photoperiod, and extraterrestrial radiation. It is also compared with three empirical models based on insolation, temperature, and a hybrid combination. In general, the greater the number of input variables, the better the performance of ML techniques, especially in combinations involving insolation that reduced the dispersion of estimated Hg on days with high atmospheric transmissivity and air temperature on days with low atmospheric transmissivity. The performance of SVM was better when compared to MLP in all statistical indicators. ML techniques presented better results than empirical models, and in general, the ordering of the best models in the three locations is achieved using SVM, MLP, and empirical models. Therefore, due to their easy implementation and generation of good results, the use of SVM models is recommended to estimate daily global radiation in the Brazilian Amazon. Full article

Internet-of-Things (IoT) approaches are continually introducing new sensors into the fields of agriculture and animal welfare. The application of multi-sensor data fusion to these domains remains a complex and open-ended challenge that defies straightforward optimization, often requiring iterative testing and refinement. To respond to this need, we have created a new open-source framework as well as a corresponding Python tool which we call the “Data Fusion Explorer (DFE)”. We demonstrated and evaluated the effectiveness of our proposed framework using four early-stage datasets from diverse disciplines, including animal/environmental tracking, agrarian monitoring, and food quality assessment. This included data across multiple common formats including single, array, and image data, as well as classification or regression and temporal or spatial distributions. We compared various pipeline schemes, such as low-level against mid-level fusion, or the placement of dimensional reduction. Based on their space and time complexities, we then highlighted how these pipelines may be used for different purposes depending on the given problem. As an example, we observed that early feature extraction reduced time and space complexity in agrarian data. Additionally, independent component analysis outperformed principal component analysis slightly in a sweet potato imaging dataset. Lastly, we benchmarked the DFE tool with respect to the Vanilla Python3 packages using our four datasets’ pipelines and observed a significant reduction, usually more than 50%, in coding requirements for users in almost every dataset, suggesting the usefulness of this package for interdisciplinary researchers in the field. Full article

Iron deficiency is a widespread public health concern, particularly in regions where rice (Oryza sativa) and beans (Phaseolus spp.) are staple foods with naturally low bioavailable iron content. Agronomic biofortification is a practical strategy to increase micronutrient levels in crops through soil, foliar, and seed-based fertilization techniques. This review synthesizes scientific studies published between 2014 and 2024 that evaluated the effectiveness of agronomic iron biofortification methods in rice and beans. The results demonstrate that site-specific interventions, such as the selection of iron sources and application methods, can improve iron concentration in grains and contribute to more nutritious and resilient food systems. However, challenges remain. There is limited information about human iron bioavailability, and the response to fertilization varies depending on soil and environmental conditions. To address these gaps, future research should include bioavailability assessments and field validation. Even so, integrating iron biofortification into standard fertilization practices is a promising approach to improve food quality and combat hidden hunger in vulnerable populations. Full article

Greenhouse gas (GHG) emissions evaluations from agroecosystems are critical, particularly as technology improves. Consistent GHG measurement methods are essential to the evaluation of GHG emissions. The objective of the study was to evaluate potential differences in gas-flux-determination (GFD) options and carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O) fluxes and growing-season-long emissions estimates from furrow-irrigated cotton (Gossypium hirsutum) in southeast Arkansas. Four GFD methods were evaluated [i.e., linear (L) or exponential (E) regression models, with negative fluxes (WNF) included in the dataset or replacing negative fluxes (RNF)] over the 2024 growing season using a LI-COR field-portable chamber and gas analyzers. Exponential regression models were influenced by abnormal CO₂ and N₂O gas concentration data points, indicating the use of caution with E models. Season-long CH₄ emissions differed (p < 0.05) between the WNF (−0.51 kg ha⁻¹ season⁻¹ for L and−0.54 kg ha⁻¹ season⁻¹ for E) and RNF (0.01 kg ha⁻¹ season⁻¹ for L and E) GFD methods, concluding that RNF options over-estimate CH₄ emissions. Gas concentration measurements following chamber closure should remain under 300 s, with one concentration measurement obtained per second. The choice of GFD method needs careful consideration to result in accurate GHG fluxes and season-long emission estimates. Full article

The degradation and loss of arable soils pose significant challenges to global food security, requiring advanced mapping and monitoring techniques to improve soil and crop management. This study evaluates the integration of Unmanned Aerial Vehicles (UAVs) and orbital sensor data for monitoring and quantifying gullies with low-cost data. The research focuses on a gully in southern Brazil, utilizing high-spatial-resolution imagery to analyze its evolution over a 25-year period (2000–2024). Photointerpretation and manual delineation procedures were adopted to define gully shoulder lines, based on low-cost and multiple-spatial-resolution data from Google Earth Pro (GEP), UAVs and conventional aerial photographs. Planimetric, volumetric, climatic, and pedological parameters were assessed and evaluated over time. Field inspections supported our interpretations. The results show that gully expansion can be effectively mapped and monitored by combining high-spatial-resolution GEP data with aerial imagery. The gully area has increased by more than 50% over the past two decades, based on GEP data, which were corroborated by submeter-resolution UAV data. The findings indicate that the erosive process remains active, progressing toward the base level. These results provide critical insights for land managers, policymakers, and agricultural stakeholders to implement targeted soil recovery strategies and mitigate further land degradation. Full article

Identifying chicken breeds and genders accurately is essential for conserving local breeds and maintaining gender ratios on farms. This study developed a system based on the Swin Transformer that efficiently and accurately classifies chicken breeds and genders. The system incorporates a target detection module to eliminate background noise and employs data augmentation techniques to prevent overfitting. A high-quality dataset, consisting of 10,482 locally captured images representing 13 Chinese native chicken breeds, was created for training and testing the model. The system was evaluated using a custom dataset and compared against popular image classification models, such as ResNet and ViT. Results indicate that the target detection module and data augmentation effectively improved the model’s performance. Additionally, strategies such as increasing the input size appropriately and utilizing pre-trained weights significantly enhanced the model’s accuracy. Interpretability analysis reveals that the system successfully identifies specific chicken body parts across different breeds and genders, aligning with human visual attention and highlighting its effectiveness. This work provides a robust solution for poultry management, aiding in tasks such as breed selection, gender ratio control, and genetic conservation. Furthermore, the methodology and dataset presented in this research provide a foundation for future studies in agricultural computer vision applications. Full article

The continuous and reliable operation of autonomous systems is important for farm management decision making, whether such systems perform crop monitoring using imaging systems or crop handling in pruning and harvesting applications using robotic manipulators. Autonomous systems, including robotic ground vehicles, drones, and tractors, are major research efforts of precision crop management. However, these systems may be less effective or require specific customizations for planting systems in low tunnels, high tunnels, or other environmentally controlled enclosures. In this work, a compact and lightweight crop management platform is developed that uses overhead power cables for continuous operation over row crops, requiring less human intervention and independent of the ground terrain conditions. The platform does not carry batteries onboard for its operation, but rather pulls power from overhead cables, which it also uses to navigate over crop rows. It is developed to be modular, with the top section consisting of mobility and power delivery and the bottom section addressing a custom task, such as incorporating additional sensors for crop monitoring or manipulators for crop handling. This prototype illustrates the infrastructure, locomotive mechanism, and sample usage of the system (crop imaging) in the application of low-tunnel-covered strawberries; however, there is potential for other row crop systems with regularly spaced support structures to adopt this platform as well. Full article

Pepper is a vital crop with extensive agricultural and industrial applications. Accurate phenotypic measurement, including plant height and stem diameter, is essential for assessing yield and quality, yet manual measurement is time-consuming and labor-intensive. This study proposes a deep learning-based phenotypic measurement method for peppers. A Pepper-mini dataset was constructed using offline augmentation. To address challenges in multi-plant growth environments, an improved YOLOX-tiny detection model incorporating a CA attention mechanism was developed, achieving a mAP of 95.16%. A detection box filtering method based on Euclidean distance was introduced to identify target plants. Further processing using HSV threshold segmentation, morphological operations, and connected component denoising enabled accurate region selection. Measurement algorithms were then applied, yielding high correlations with true values: $R^2$ = 0.973 for plant height and $R^2$ = 0.842 for stem diameter, with average errors of 0.443 cm and 0.0765 mm, respectively. This approach demonstrates a robust and efficient solution for automated phenotypic analysis in pepper cultivation. Full article

Strawberry harvesting represents a labor-intensive agricultural operation where existing end-effector pose control algorithms frequently exhibit insufficient precision in fruit grasping, often resulting in unintended damage to target fruits. Concurrently, deep learning-based pose control algorithms suffer from inherent training instability, slow convergence rates, and inefficient learning processes in complex environments characterized by high-density fruit clusters and occluded picking scenarios. To address these challenges, this paper proposes an enhanced reinforcement learning framework DRFW-TQC that integrates Dynamic L2 Regularization for adaptive model stabilization and a Group-Wise Feature Weighting Network for discriminative feature representation. The methodology further incorporates a picking posture traction mechanism to optimize end-effector orientation control. The experimental results demonstrate the superior performance of DRFW-TQC compared to the baseline. The proposed approach achieves a 16.0% higher picking success rate and a 20.3% reduction in angular error with four target strawberries. Most notably, the framework’s transfer strategy effectively addresses the efficiency challenge in complex environments, maintaining an 89.1% success rate in eight-strawberry while reducing the timeout count by 60.2% compared to non-adaptive methods. These results confirm that DRFW-TQC successfully resolves the tripartite challenge of operational precision, training stability, and environmental adaptability in robotic fruit harvesting systems. Full article

Precision agriculture increasingly relies on real-time data from soil sensors to optimize irrigation and nutrient application. Soil moisture and electrical conductivity (EC) are key indicators in irrigation and fertigation systems, directly affecting water-use efficiency and nutrient delivery to crops. This study evaluates the performance of an IoT-based soil-monitoring system for real-time tracking of EC and soil moisture under varied fertigation conditions in both laboratory and field scenarios. The EC sensor showed strong agreement with laboratory YSI measurements (R² = 0.999), confirming its accuracy. Column experiments were conducted in three soil types (sand, sandy loam, and loamy sand) to assess the EC and soil moisture response to fertigation. Sand showed rapid infiltration and low retention, with EC peaking at 420 µS/cm and moisture 0.33 cm³/cm³, indicating high leaching risk. Sandy loam retained the most moisture (0.35 cm³/cm³) and showed the highest EC (550 µS/cm), while loamy sand exhibited intermediate behavior. Fertilizer-specific responses showed higher EC in Calcium Ammonium Nitrate (CAN)-treated soils, while Monoammonium Phosphate (MAP) showed lower, more stable EC due to limited phosphorus mobility. Field validation confirmed that the IoT system effectively captured irrigation and fertigation events through synchronized EC and moisture peaks. These findings highlight the efficacy of IoT-based sensor networks for continuous, high-resolution soil monitoring and their potential to support precision fertigation strategies, enhancing nutrient-use efficiency while minimizing environmental losses. Full article

Accurate estimation of the leaf area index (LAI), a key indicator of canopy development and light interception, is essential for improving productivity in greenhouse tomato cultivation. This study presents a non-destructive LAI estimation method using side-view images captured by a vertical scanning system. The system recorded the full vertical profile of tomato plants grown under two deleafing strategies: modifying leaf height (LH) and altering leaf density (LD). Vegetative and leaf areas were extracted using color-based masking and semantic segmentation with the Segment Anything Model (SAM), a general-purpose deep learning tool. Regression models based on leaf or all vegetative pixel counts showed strong correlations with destructively measured LAI, particularly under LH conditions (R² > 0.85; mean absolute percentage error ≈ 16%). Under LD conditions, accuracy was slightly lower due to occlusion and leaf orientation. Compared with prior 3D-based methods, the proposed 2D approach achieved comparable accuracy while maintaining low cost and a labor-efficient design. However, the system has not been tested in real production, and its generalizability across cultivars, environments, and growth stages remains unverified. This proof-of-concept study highlights the potential of side-view imaging for LAI monitoring and calls for further validation and integration of leaf count estimation. Full article

Agriculture in India supports millions of livelihoods and is a major force behind economic expansion. Challenges in modern agriculture depend on environmental factors (such as soil quality and climate variability) and biotic factors (such as pests and diseases). These challenges can be addressed by advancements in technology (such as sensors, internet of things, communication, etc.) and data-driven approaches (such as machine learning (ML) and deep learning (DL)), which can help with crop yield and sustainability in agriculture. This study introduces an innovative deep neural network (DNN) approach for identifying leaf diseases in paddy crops at an early stage. The proposed neural network is a hybrid DL model comprising feature extraction, channel attention, inception with residual, and classification blocks. Channel attention and inception with residual help extract comprehensive information about the crops and potential diseases. The classification module uses softmax to obtain the score for different classes. The importance of each block is analyzed via an ablation study. To understand the feature extraction ability of the modules, extracted features at different stages are fed to the SVM classifier to obtain the classification accuracy. This technique was experimented on eight classes with 7857 paddy crop images, which were obtained from local paddy fields and freely available open sources. The classification performance of the proposed technique is evaluated according to accuracy, sensitivity, specificity, F1 score, MCC, area under curve (AUC), and receiver operating characteristic (ROC). The model was fine-tuned by setting the hyperparameters (such as batch size, learning rate, optimizer, epoch, and train and test ratio). Training, validation, and testing accuracies of 99.91%, 99.87%, and 99.49%, respectively, were obtained for 20 epochs with a learning rate of 0.001 and sgdm optimizer. The proposed network robustness was studied via an ablation study and with noisy data. The model’s classification performance was evaluated for other agricultural data (such as mango, maize, and wheat diseases). These research outcomes can empower farmers with smarter agricultural practices and contribute to economic growth. Full article

The most common size measurements for agricultural produce, including fruits and vegetables, are length and width. While the length of any agricultural produce can be unique, the width varies continuously along its length. Single-width measurements alone are insufficient for accurately characterizing varying width profiles, resulting in an inaccurate representation of the shape or mean dimension. Consequently, the manual measurement of multiple mean dimensions is laborious or impractical, and no information in this domain is available. Therefore, an efficient alternative computer vision measurement tool was developed utilizing ImageJ (Ver. 1.54p). Twenty sample sets, comprising fruits and vegetables, with each representing different shapes, were selected and measured for length and multiple widths. A statistically significant minimum number of multiple widths was determined for practical measurements based on an object’s shape. The “aspect ratio” (width/length) was identified to serve as an effective indicator of the minimum multiple width measurements. In general, 50 multiple width measurements are recommended; however, even 15 measurements would be satisfactory ($\mathrm{1.0\%}\pm \mathrm{0.6\%}$ deviation from 50 widths). The developed plugin was fast (734 ms ± 365 ms CPU time/image), accurate (>99.6%), and cost-effective, and it incorporated several user-friendly and helpful features. This study’s outcomes have practical applications in the characterization, quality control, grading and sorting, and pricing determination of agricultural produce. Full article

Stones in agricultural fields pose a recurring challenge, particularly due to their potential to damage agricultural machinery and disrupt field operations. As modern agriculture moves toward automation and precision farming, efficient stone detection has become a critical concern. This study explores the potential of thermal imaging as a non-invasive method for detecting stones under varying environmental conditions. A series of controlled laboratory experiments and field investigations confirmed the assumption that stones exhibit higher surface temperatures than the surrounding soil, especially when soil moisture is high and air temperatures are cooling rapidly. This temperature difference is attributed to the higher thermal inertia of stones, which allows them to absorb and retain heat longer than soil, as well as to the evaporative cooling from moist soil. These findings demonstrate the viability of thermal cameras as a tool for stone detection in precision farming. Incorporating this technology with GPS mapping enables the generation of accurate location data, facilitating targeted stone removal and reducing equipment damage. This approach aligns with the goals of sustainable agricultural engineering by supporting field automation, minimizing mechanical inefficiencies, and promoting data-driven decisions. Thermal imaging thereby contributes to the evolution of next-generation agricultural systems. Full article

This study presents a bibliometric review aimed at mapping and analyzing the scientific literature related to the ethical implications of artificial intelligence (AI) in livestock farming, which is a rapidly emerging yet still underexplored field in international research. Based on the Scopus database, 151 documents published between 2015 and 2025 were identified and analyzed using the VOSviewer version 1.6.20 and Biblioshiny for Bibliometrix (RStudio version 2023.12.1) tools. The results show a significant increase in publications from 2021 onwards, reflecting the growing maturity of discussions around the integration of digital technologies in the agricultural sector. Keyword co-occurrence and bibliographic coupling analyses revealed the formation of four main thematic clusters, covering technical applications in precision livestock farming as well as reflections on governance, animal welfare, and algorithmic justice. The most influential authors, high-impact journals, and leading countries in the field were also identified. As a key contribution, this study highlights the lack of robust ethical guidelines and proposes future research directions for the development of regulatory frameworks, codes of conduct, and interdisciplinary approaches. The findings underscore the importance of aligning technological innovation with ethical responsibility and social inclusion in the transition to digital livestock farming. Full article

Multispectral images provide valuable indicators of crop nutritional status, playing a key role in strategies to reduce fertilizer use and enable supplementary applications in cases of nitrogen deficiency, thereby ensuring productivity and profitability for farmers. However, the diversity of remote sensing platforms (RSPs) makes the choice challenging, as there are few comparative studies. This study compares the remote sensing platforms Sentinel-2, CBERS-04A, and unmanned aerial vehicle (UAV), assessing their accuracy in detecting different nitrogen doses (NDs) throughout the maize crop cycle in Botucatu-SP, using 10 vegetation indices (VIs). Six NDs were tested (0, 36, 84, 132, 180, and 228 kg ha⁻¹ of nitrogen) in nine assessments during the crop cycle. The results showed that, at the V7 stage, the RSPs were effective in detecting the NDs in eight VIs. However, at the VT stage, only the Sentinel-2 and CBERS-04A satellites demonstrated effectiveness in six VIs. Despite the high correlation among the RSPs, the ability to distinguish the NDs varied depending on the vegetation index (VI) and phenological stage. These findings highlight the importance of selecting the appropriate VI and optimal timing, regardless of the chosen platform. Full article

Prays oleae Bernard, 1788, or the olive moth, is a significant pest in Croatian olive groves. This study aims to develop a functional model based on an artificial neural network to detect olive moths in real time. This study was conducted in two different orchards in Zadar County, Croatia, in the periods from April to September 2022 and from May to July 2023. Moth samples were collected by placing traps with adhesive pads in these orchards. Photos of the pads were taken every week and were later annotated and used to develop the dataset for the artificial neural network. This study primarily focused on the average precision parameter to evaluate the model’s detection capabilities. The average AP value for all classes was 0.48, while the average AP value for the Olive_trap_moth class, which detected adult P. oleae, was 0.59. The model showed the best results at an IoU threshold of 50%, achieving an AP50 value of 0.75. The AP75 value was 0.56 at an IoU = 75%. The mean average precision (mAP) was 0.48. This model is a promising tool for P. oleae detection; however, further research is advised. Full article

This study investigates the impact of biogas slurry ratio, hole diameter and depth under hole irrigation on the soil wetting front migration distance and cumulative infiltration. In this study, a model describing the water transport characteristics of biogas slurry hole irrigation was developed based on the HYDRUS model. Results demonstrated that the HYDRUS model can be used for biogas slurry hole irrigation (NSE > 0.952, PBIAS ≤ ±0.34). Furthermore, the study revealed that the soil cumulative infiltration and soil wetting front migration distance decreased gradually with an increase in the biogas slurry ratio, while they increased gradually with an increase in the hole diameter and depth. The lateral and vertical wetting front migration distances exhibited a well-defined power function relationship with the soil’s stable infiltration rate and infiltration time (R² ≥ 0.977). The soil wetting front migration distance curve can be represented by an elliptic curve equation (R² ≥ 0.957). Additionally, there was a linear relationship between the cumulative infiltration and soil wetted body area (R² ≥ 0.972). Soil wetting front migration distance model ($X=4.442f_0^{0.375}{t}^{0.24}, Z=11.988f_0^{0.287}{t}^{0.124}, f_0=96.947K_s^{1.151}{D}^{0.236}{H}^{1.042}$, NSE > 0.976, PBIAS ≤ ±0.13) and cumulative infiltration model ($I=0.3365S$, NSE > 0.982, PBIAS ≤ ±0.10) established under biogas slurry hole irrigation exhibited good reliability. This study aims to determine optimal hole diameter, depth, and irrigation volume for biogas slurry hole irrigation by establishing a model for soil wetting front migration distance and cumulative infiltration based on crop root growth patterns, thereby providing a scientific basis for its practical application. Full article