Proceedings
Topics
| Filter results50 paper(s) found. |
|---|
1. Integrating Proximal Hyperspectral and Machine Learning to Predict Nitrogen in Short- and Full-stature Corn Hybrids at Early Growth Stage in Indiana, USANitrogen (N) fertilizer use is a complex challenge, as underapplication can harm yield and overapplication can harm profitability and the environment. N accounts for roughly 58% of total US corn fertilizer use (an annual expense of ~$8 billion), with overapplication estimated at 15% ($1.2 billion for possible savings). Within this setting, early-season yield prediction is a high-value capability for breeding and farmers. If plot-level plant N can be forecasted with high accuracy before the co... B. Paulus scheffer, D. . Quinn, P.H. Magalhaes cisdeli, J. Jin, Z. Qin, I. Ciampitti |
2. Early Forecasting of Maize Lodging Risk Through Multi-period and Multi-source Data IntegrationLodging is a critical constraint on global maize (Zea mays L.) productivity, primarily through detrimental effects on both grain yield and quality. However, reliable methods to predict maize lodging risk early in the growing season are lacking, which hinders timely implementation of effective agronomic management interventions to increase crop lodging resistance and reduce corresponding yield losses. This work aimed to develop a feasible early season maize lodging risk prediction met... L. Dong, Y. Miao, X. Wang, P. Berry, D. Hatley, K. Kusnierek |
3. Spatial Prediction of Soil Classes and Nutrients Using Random Forest in the Context of Precision ViticulturePrecision viticulture is based on modeling the spatial variability of soil, plant, and topographic attributes to support optimized management decisions. In this context, machine learning based spatial prediction algorithms have been increasingly applied for spatial interpolation. Their application in vineyards has shown strong potential to improve the representation of spatial variability and to support site-specific management strategies in viticulture. The objective of this study was to eva... F. Lasch, B. Trevizan paese, J.M. Moura-bueno, G. Brunetto , A.A. Kokkonen, F. De araújo pedron, R.S. Dalmolin, L. De paula amaral |
4. Large-Scale Sugarcane Yield Prediction Across Regions by Integrating Multi-Source Remote Sensing and Machine LearningSugarcane (Saccharum officinarum L.) is one of the most important agro-industrial crops worldwide, playing a key role in sugar, bioethanol, and renewable energy production. Early and accurate yield estimation during the growing season is essential to support agricultural planning, resource management, and decision-making in the sugar-energy industry under increasing climate variability. However, most yield models are calibrated to single locations and struggle to transfer across regions. The ... R. Fortes gallego, F. Serra burriel, M. Cabrera dengra, C. Ferraz, A. Do vale dondo |
5. Precision Phenotyping for Yield Prediction in SoybeanIn the contemporary landscape of Precision Agriculture 4.0, the rapid and non-destructive quantification of plant structural traits is a cornerstone for accelerating soybean breeding programs and optimizing field-level cultivation strategies. Traditional manual phenotyping methodologies are inherently limited by high labor intensity, significant subjectivity, and low throughput, which hinder the ability to analyze large populations during critical reproductive stages. To bridge this gap,... W. Su |
6. Democratizing Prescriptive Agronomy: Quality-Preserving Edge AI for Sugar BeetsThe global sugar beet sector faces a critical production paradox where agronomic interventions designed to maximize root yield often compromise sucrose concentration and processing quality. While precision agriculture aims to navigate this delicate balance, current methodologies have reached a methodological impasse. Existing solutions are bifurcated between descriptive data-intensive machine learning (ML), which struggles to generalize across heterogeneous fields, and physiological Process-B... A. Tabbassi, S. Henkler |
7. Can Management Zones Be Useful in Guiding Mechanized Peanut Harvesting?Mechanized peanut harvesting presents challenges due to the crop’s indeterminate growth habit, which results in non-uniform maturation across the field; the ideal harvest point is reached when the maturity index exceeds 0.7. Consequently, analyzing the spatial variability of maturation is essential for identifying homogeneous areas and guiding harvest at the optimal time. In this context, Management Zones (MZs), as a Precision Agriculture tool, enable the subdivision of fields into more... A. Andrade da silva, T.C. Moura oliveira, E. Sales, S. Luns, E.M. Perussi, R.H. De souza silva, S. Luns, R.P. Silva, J. , A.L. De brito filho , R.P. Silva |
8. A Decision Support Tool for Developing Aflatoxin Risk Maps in Peanut FieldsAspergillus flavus (A. flavus) is a soil fungus that contaminates preharvest peanuts (Arachis hypogea) with the carcinogenic secondary metabolite aflatoxin. Because aflatoxin can cause serious illness or death at low concentrations (μg kg-1 to mg kg-1), its presence in foods and feeds is strictly regulated by food safety agencies around the world. Based on previous research by the authors, a hypothesis was developed that aflatoxin ... G. Vellidis, S. Maktabi, K. Boote, G. Hoogenboom, L. Lacerda, C. Pilcon, S. Shrestha, R. Wiggins |
9. High-resolution Orbital Imagery and Neural Networks to Predict Brix and Purity in SugarcaneIntegrating artificial neural networks with high-resolution satellite remote sensing data can provide non-destructive indicators for assessing sugarcane quality at field scale. Conventional laboratory methods for sucrose-related quality assessment are costly, labor-intensive, and operationally demanding, particularly when applied continuously over large commercial areas. This study evaluated the potential of multispectral imagery from the PlanetScope CubeSat platform, vegetation indices, and ... P. Cardoso, R.P. Silva, T.R. Da silva, M.F. De oliveira, J.B. Souza, S.L. De almeida |
10. Use of Textural and Spectral Data in Predictive Modeling of Sugarcane YieldSugarcane is one of the most important crops in Brazil, playing a strategic role in the production of sugar, ethanol, and bioenergy. Efficient monitoring of crop yield is essential for agricultural management and decision-making; however, conventional yield estimation methods are generally labor-intensive, destructive, and inefficient in capturing spatial variability within fields. In this context, the use of remote sensing techniques integrated with machine learning models emerges as a promi... L. Rodrigues , S. Luns, G. Rolim, T. Canata, V. Carreira |
11. Advanced Method to Assess the Impact of Soil Fertility Variability on Maize (Zea mays L.) Productivity.Maize production in Brazil for the 2024/2025 season is estimated at 119.6 million tons, representing a 3.4% increase compared to the previous cycle, despite a 5.9% reduction in cultivated area. This scenario highlights the need for more efficient agricultural systems capable of sustaining high productivity levels under land-use restrictions. Traditionally, high-yield agriculture has relied on intensive input use, a practice that can lead to waste. In contrast, precision agriculture emerges as... D. Portioli sampaio, J.M. Villela, P.E. Cruvinel |
12. Smartphone-Based RGB Phenotyping of Hydroponic Lettuce Growth DynamicsHydroponic lettuce production requires frequent, accurate growth assessment to optimize yield, nutrient efficiency, and crop uniformity within short, nutrient‑sensitive cycles. Existing imaging systems can deliver such precision but are often costly or technically demanding, limiting adoption in smaller hydroponic operations. Smartphone‑based imaging offers a practical alternative for scalable proximal phenotyping, yet its cross‑device quantitative accuracy and performance under operati... L. Katz, N. Snir, K. Genkin, N. Rotbart, E. Nevo, N. Ronen, O. Reichmann |
13. Estimating Grape Bunch Yield Using Convolutional Neural Networks and Proximal RGB Imaging in the Brazilian PampaViticulture of fine wines has become an increasingly important economic activity in the Pampa biome of southern Brazil, a relatively recent production frontier with approximately two decades of commercial development. In this emerging region, accurate prediction of grapewine productivity represents one of the most relevant challenges for growers, as reliable early estimates directly support decision-making related to harvest planning, logistics, labor allocation, and marke... S. Camargo, E.M. Da silveira, F.I. Nogueira, A.F. Campos, V.Z. Mércio |
14. Predicting Peanut Maturity and Yield by Integrating Multicriteria Regression and Remote Sensing.Accurate predictions of yield and maturity are essential for optimizing crop management in low-tech crops such as peanuts. However, the spatial and temporal variability of these variables poses significant challenges for conventional methods. Yield can be determined using labor-intensive manual methods or sensors on harvesters. Despite advances, there is a scarcity of studies that simultaneously integrate yield and maturity prediction for peanuts, a critical step for harvest optimization. The... J. Souza, J. Lucas da silva ferreira , S. Luns, M.F. Oliveira, W. Sousa |
15. GAIG: High‑Resolution Wall‑to‑Wall Modelling of Within‑Field Spatial Variability in Crop YieldQuantifying the temporal stability and causes of within‑field variation in crop yield is fundamental to precision‑agriculture research, particularly as agricultural areas seek to identify lands with persistently low productivity that may constitute marginal cropland. What is needed is a modelling framework capable of using spatial patterns in yield to reveal stability zones, diagnose sources of variability, and enable consistent comparison across farms and years. Accordingly, the objectiv... |
16. Integrating Tractor-tire-tool Adjustable Parameters and UAV‑derived Soil Indices to Predict Fuel Consumption and Crop Emergence in Spring Barley SowingThe optimization of energy use and agronomic performance in agricultural operations has become a central challenge in modern agriculture. To achieve this dual objective, farmers could adjust the machinery settings of a tractor-tire-tool system to ensure efficient resource utilization while maintaining optimal agronomic outcomes. This study was conducted as a part of the AgrEnOp project, which aims to predict fuel consumption (l/ha) and crop emergence (plant/m2) based on operator-ad... D. Urbina salazar, A. Yatskul, F. Pinet, A. Dujany, C. Ugarte |
17. Multivariate Analysis of Structural, Climatic, and Management Factors Associated with Cotton Yield in the Brazilian MidwestUnderstanding the integrated drivers of cotton yield in highly intensified production systems remains a central challenge for precision agriculture. This study aimed to assess, from a systemic perspective, the structural, climatic, phytosanitary, and management factors associated with cotton yield during the 2024/25 season in large-scale commercial farms located in the states of Mato Grosso and Mato Grosso do Sul, Brazil. Approximately 91,000 ha were analyzed across multiple farms and... R. Rimoldi tavanti, G. Morais, D. |
18. Development of a Predictive Machine Learning Model for Pasture Biomass Using Satellite Vegetation Indices and Climate Data in Spanish Dehesa SystemsExtensive livestock systems are fundamental to the ecological, economic, and cultural sustainability of Mediterranean agroecosystems such as the Spanish dehesa. These silvopastoral landscapes support biodiversity, prevent land abandonment, and sustain rural livelihoods, but their productivity is highly dependent on pasture availability. Efficient management therefore requires accurate and timely information on pasture biomass, which is strongly influenced by climatic variability, soil propert... C. Ferraz, A. Tamayo lópez, A. Do vale dondo |
19. Quantifying Prediction Uncertainty in Field-scale Soil Maps Generated by Machine Learning.Field-scale maps of soil properties are a key component of precision agriculture, as they are routinely used as inputs for variable-rate fertilization, zone delineation, and site-specific management. While machine learning models have substantially improved the accuracy of spatial predictions, uncertainty associated with these predictions is often ignored, limiting the reliability of soil maps as decision-support tools. Quantifying prediction uncertainty is essential not only to assess map q... F. García seleme, P. Paccioretti, M. Balzarini, M. Córdoba |
20. Generating Data Via Operational Monitoring of Backpack Equipment on Small FarmsThe Brazilian coffee industry, a global leader in production and exports, faces the challenge of increasing production efficiency to meet growing worldwide demand while preserving natural resources. Precision Agriculture (PA) offers essential tools for this data-driven sustainable intensification; however, its adoption in regions with rugged topography and by family-based growers is severely limited by the scarcity of accessible technologies, particularly for spatial yield measurement. The se... V. Ferraz, L. Mariotto nabarro, R. Castanho fernandes, B.B. Barreto, B. Ricardo silva costa, J.P. Molin |
21. Influence of Terrain Attributes on the Spatial Variability of Soil Macronutrients in Vineyards of Southern BrazilNutrient variability in vineyards directly affects grapevine development and grape yield, highlighting the importance of appropriate nutritional management, since optimizing soil nutrient levels contributes to improved grape, and, consequently, wine quality. The objective of this study was to evaluate the spatial variability of soil macronutrient distribution in a vineyard and to correlate it with terrain attributes. The study was conducted in a 10-ha commercial Pinot Noir vineyard located in... A. Costa tolfo, B. Trevizan paese, J.M. Moura bueno, A.A. Kokkonen, G. Brunetto, S. Schemmer |
22. Relationship Between Soil Classes and Grape Yield in a Vineyard of the Campanha Gaúcha RegionBrazilian viticulture has shown significant expansion in recent decades. However, this productive growth has brought new challenges for vineyard management, particularly regarding the understanding of soil spatial variability and its relationship with grape yield and quality. The objective of this study was to correlate the spatial variability of soil classes with grape yield. The study was conducted in a commercial vineyard located in Santana do Livramento, in the Campanha Gaúcha regi... B. Baumgardt, B. Trevizan paese, J.M. Moura bueno, G. Brunetto, A.A. Kokkonen, A. Benetti |
23. Harmonic Modeling of Coffee Biennial Bearing to Quantify Between-plot Variability: a Precision Agriculture Approach for Small-scale AgriculturePrecision Agriculture (PA) practices rely on detecting spatiotemporal variability within a plot to delineate subplots by pixels or by management zones (MZs). This paradigm has been assumed for large-scale plots, yet their adoption in smallholder systems, such as coffee crops under family-based agriculture, remains limited. In this context, within-plot variability is often less operationally relevant than between-plots divergency. Therefore, we assume each plot as a MZ and focus on manage them... J.P. Molin, B. Costa, B. Barreto |
24. A Methodological Framework for Modeling Plant Virus Occurrence Using Biometeorological Data: Insights from Multi-crop Case Studies in ArgentinaViral diseases represent a major threat to the productive stability of agricultural systems. Their spatial and temporal occurrence is influenced by environmental conditions that regulate interactions among viruses, vectors, and hosts, making disease anticipation difficult using statistical traditional approaches. This situation highlights the need to understand the dynamics of the different biological components capable of affecting agricultural systems, and design and apply tools that facili... F. Suarez, B. Gómez montenegro, C. Dottori, V. Alemandri, S. De breuil, C. Bruno, F. García seleme |
25. Evaluating APSIM for Precision Optimization of Planting Windows and Nitrogen Management in Maize-Soybean Intercropping Systems in MalawiMaize-soybean intercropping is a key strategy for improving food security and resource-use efficiency in smallholder systems in sub-Saharan Africa. In Malawi, soybean promotion supports sustainable intensification, yet optimizing planting windows, spatial arrangements, and nitrogen (N) management under variable rainfall remains challenging. This study assesed the capability of the Agricultural Production Systems Simulator (APSIM) to simulate maize-soybean performance and identify precis... |
26. Simulation of Different Nitrogen Fertilization Strategies Using Management Zones in Sugarcane CultivationNitrogen plays a central role in sugarcane physiology, as it is a structural component of amino acids, proteins, nucleic acids, and chlorophyll, being essential for photosynthetic activity, leaf expansion, biomass accumulation, and stalk formation. Given its relevance, nitrogen management efficiency represents a central research topic in sugarcane, particularly in systems characterized by strong spatial heterogeneity, where soil physical–hydric attributes, fertility levels, and environm... G.V. Bedum, J.P. Molin, R. Canal filho |
27. Upscaling UAV Image-Trained Machine Learning Models from Research Plots to Commercially Cropped LandHigh-throughput plant phenotyping (HTPP) leverages the advancement of unmanned aerial vehicles (UAVs) technology, paired with improvement in spectral sensing technology to allow for the derivation of plant phenotypic traits from image analysis. Crop breeding programs continue to increase incorporation of HTTP methods into their pipelines to enhance their efficiency of selecting for varieties. Machine learning (ML) models, often used hand in hand with HTTP methods, generate phenotypic trait pr... W. Maess, S. Shirtliffe, K. Nketia |
28. Study of Temporal Microclimatic Variability and Its Impact on Soybean (Glycine max (L.) Merrill) Development in a Protected Environment: A Precision Agriculture ApproachSoybean, belonging to the Fabaceae family, is a leguminous crop of high economic and nutritional value, widely cultivated in Brazil, which ranks among the world’s largest producers and exporters. However, the growing demand for productivity has driven the adoption of digital technologies and Precision Agriculture (PA) methods aimed at intelligent crop management. In protected environments, cultivation allows the mitigation of external constraints, although it does not eliminate internal... J. Risardi, P. Herrmann junior, A. Torre neto, P. Cruvinel |
29. Comparison of Machine Learning Models for Within-field Cotton Yield Prediction: Assessing the Value of Temporal Data and GeneralizabilityCotton (Gossypium hirsitum L.) yield prediction is challenging due to the complex interactions between static environmental factors, such as soil properties and topography, and dynamic variables, including weather and crop management. These factors generate significant spatial and temporal variability within and across fields. While remote sensing technologies, particularly satellite-derived vegetation indices, provide spatially explicit data to evaluate crop health, translating... L. Bastos, L. Fuhrer, W. Porter, G.J. Scarpin, A. Kaur dhaliwal, A. Bhattarai, A. Jakhar |
30. Bridging Continents for Soil Carbon Mapping: A Transfer Learning Framework from European LUCAS to Chinese FarmlandsSoil organic carbon (SOC) mapping is fundamental to precision agriculture and climate change mitigation. Building accurate SOC prediction models typically requires extensive local sampling, which is costly and time-consuming. Can spectral-SOC relationships learned from large-scale soil databases be transferred across continents? This study addresses this question by developing a transfer learning framework that leverages the European LUCAS database (n=2288) to predict SOC in Chinese farmlands... X. Chen, Q. Cao |
31. Soil Water Nowcasting for Site-specific Yield Potential EstimationKnowing how much plant available water (PAW) is stored across a field at key decision points in the growing season is fundamental to precision agriculture. Spatial variability in soil water translates directly into variability in water-limited yield potential, yet most growers lack the tools to quantify this at the within-field scale. Here we present a Soil Water-Energy Balance (SWEB) model that offers a framework to deliver daily, 30 m resolution estimates of PAW across any dryland padd... T. Bishop, Y. Yu, M.J. Tilse, P. Filippi |
32. An Integrated Water–Energy Vulnerability Index for Irrigated Agricultural RegionsThe growing interdependence between water availability and energy infrastructure has significantly increased the vulnerability of irrigated agricultural regions, particularly under conditions of climate variability, hydrological uncertainty, and seasonal demand peaks. Irrigated production systems simultaneously depend on reliable water supply and stable energy provision, making them particularly sensitive to disruptions in either domain. Although the water–energy nexus literature has ad... T.A. Rodolfo, P.S. Schneider, M.A. Perez, F.D. Mantovan, H.R. Bressan, A.C. Reginatto |
33. Operational Satellite Weed Detection Across 14,000 Sugarcane Fields: Lessons in Temporal Feature DesignWeed infestations in sugarcane (Saccharum officinarum L.) can reduce yields by 20–60% depending on species composition and management timing, yet operational weed management at scale remains an unsolved challenge. Existing studies typically cover tens of fields; scaling to thousands introduces challenges in processing throughput, ground truth scarcity, and feature design. This work describes the development and operational deployment of a satellite-based weed detection system ... C. Ferraz, R. Fortes, M. Cabrera dengra, J. Poli, E. Bernardes júnior, A. Do vale dondo |
34. Early Yield Estimation in Hass Avocado Using Ecophysiological Variables and Machine LearningThis study evaluated the ability of machine learning models to estimate yield in mature Hass avocado trees (>5 years), using ecophysiological variables measured with MultispeQ v2.0 (RIDES 2.1 protocol) and electrical capacitance (1 Hz). The study was conducted at Pan de Azúcar farm (Villahermosa, Tolima, Colombia; 1,565 m a.s.l., Andisols) on 60 trees, with data collected across four phenological stages (fruit development, fruit maturation, leaf and shoot growth, and pre-flowering) ... D. Rayo Álvarez, P.J. Murillo sandoval, A.E. Darghan contreras, D.F. Conejo rodriguez |
35. Temporal NDRE Dynamics from UAS Imagery to Characterize Rice Drought ResponseCharacterizing drought resilience in rice remains challenging under increasing climate variability. Drought tolerance is a complex and dynamic trait that is difficult to quantify using traditional field phenotyping approaches, particularly when responses vary with time. High-throughput temporal phenotyping with unmanned aircraft systems (UAS) enables monitoring of canopy reflectance dynamics associated with water stress across the growing season. This study evaluated whether ... |
36. Advanced 2D and 3D Image-Based Plant Phenotyping of Citrus Morphological Responses to Candidatus Liberibacter Asiaticus InfectionHuanglongbing (HLB), associated with Candidatus Liberibacter asiaticus (CLas), is the most destructive citrus disease worldwide and threatens the long-term sustainability of production because all commercially cultivated varieties are susceptible. Identifying tolerant or resistant genotypes has therefore become a central priority for modern citrus breeding programs. Conventional phenotyping based on visual scoring and manual measurements is time-consuming, labor intensive, s... J. Cifuentes arenas, C. Lunewski, F. Keil, M.N. Alves, N.A. Wulff |
37. Plant-Level Coffee Production Estimation Based on Morphological IndicesProduction estimation in coffee farming is traditionally conducted at aggregated spatial scales, which often limits the characterization of variability among individual plants and constrains its applicability for precision-oriented management. In production systems where within-field heterogeneity affects decisions related to harvesting, logistics, and crop management, approaches capable of representing plant-level variability become particularly relevant. Within this context, this study prop... D. Queiroz, D.H. Leite, D. Sárvio valente, G. Dumbá monteiro de castro, D.B. Marin |
38. Field-scale Prediction of Soil Organic Carbon Using Integrated Proximal Sensing and Terrain CovariatesThe knowledge of soil organic carbon (SOC) is essential for climate change mitigation strategies, soil security, and management within precision agriculture scenarios in agricultural areas. The use of approaches integrating spectral and magnetic sensor data with topographic covariates has shown promise for predicting SOC along the soil profile. In this context, the study aimed to develop predictive models of SOC content at depth through the integration of proximal sensing data and topographic... J. Moura bueno, L.F. rech, R.S. Diniz dalmolin, L. De paula amaral, I. Buana, F. De araujo pedron |
39. Integration of Spectrotemporal Metrics and Machine Learning for Soybean Grain Yield PredictionEstimating agricultural grain yield in heterogeneous production environments remains one of the main challenges of digital agriculture. Although vegetation indices, such as the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Red Edge (NDRE), are widely used to describe canopy vigor, their predictive capacity strongly depends on how temporal information is represented and on the structure of the model employed to integrate this spectral variability. The present stud... L. Rossetto gerlach, A.L. Vian, C. Bredemeier, T. Enderle, T. Santos cocco, M. Wrubleski |
40. Predicting Yield Stability Classes Using Satellite Imagery in the Absence of Yield Monitor DataSite-specific management is essential for improving agricultural productivity while reducing input costs and minimizing environmental impacts. Although yield monitor data are commonly used to characterize within-field yield variability, their availability is often limited by technological and economic constraints. The primary objective of this study was to compare spatial–temporal stability classes derived from yield monitor data and satellite imagery in cotton produ... K. Poudel, A. Bhattarai, A. Jakhar, L. Bastos, A. Dhaliwal |
41. Main Environmental and Variety Drivers of Cotton Seed Quality: Historical Insights from the United States Cotton BeltCotton seed quality traits including oil content, nitrogen (protein), and gossypol significantly influence seed value and end-use applications, yet their predictability based on environmental conditions across varied U.S. growing regions remains poorly understood. This study aimed to: (i) identify critical environmental predictors of seed composition; (ii) build machine learning models to predict seed quality as a function of seasonal weather patterns; and (iii) assess dif... A. Dhaliwal, L. Bastos, K. Sv, A. Bhattarai, A. Jakhar, K. Poudel, D.M. Mccallister, S.Y. Jaconis |
42. Estimation of carbon sequestration in agricultural crops using the C-Questro softwareContemporary agriculture faces the challenge of reconciling productivity with climate impact mitigation, positioning soil and plant biomass carbon sequestration as a strategic pillar for global sustainability. However, carbon quantification at the field scale still encounters hurdles due to high-cost methodologies or operational complexity. The objective of this work was to develop and validate a Python-based software, named "C-Questro," designed to automate the estimation of carbon... |
43. Climatic Zoning of the Peanut Cercosporiosis Complex in São Paulo Under Climate Change ScenariosThe cercosporiosis complex is an important foliar disease of peanut, caused by the fungi Cercospora arachidicola and Nothopassalora personata, impacting grain yield and quality. Another relevant aspect is the symptoms of defoliation and vegetative weakening caused by these fungi, which may lead to significant losses during the digging and harvesting stages of peanut, a crop intrinsically associated with mechanization. The objective of this study was to develop a climatic zoning of the cercosp... R. Mendes, I. De oliveira vieira, R.P. Silva |
44. Predicting Pre-harvest Cotton Fiber Quality: An Open-data and Machine Learning FrameworkIntra-field variability in soil properties and topography, and inter-field variability in weather patterns often leads to inconsistent cotton fiber quality and yield outcomes, posing challenges for growers. This study aims to: (i) predict within-field cotton fiber quality traits based on environmental and soil variables using machine learning models; (ii) identify the most influential environmental drivers (weather, vegetation indices, soil properties, terrain characteristics) affecting cotto... |
45. Development and Validation of a Low-Cost IoT-Based Weather Station Using LoRa Communication for Precision AgricultureAccess to accurate local meteorological data remains a critical bottleneck for precision agriculture adoption among small and medium-scale Brazilian farmers. Commercial weather stations cost between R$ 15,000 and R$ 50,000, while public networks such as INMET operate with average inter-station spacing of 30–50 km, insufficient to capture the microclimate variability that drives field-scale irrigation and crop management decisions. This study presents the development, field validation, a... A.L. Carvalho, C.C. Santana, R. Avelar, F.D. Silva, F. Soares |
46. Integration of LiDAR-Derived TWI and UAV Multispectral Data for Waterlogging Susceptibility Mapping in Precision AgricultureTopography directly controls water redistribution across the landscape, shaping the spatial variability of soil moisture in agricultural areas. The Topographic Wetness Index (TWI), derived from digital elevation models, is widely used to estimate the potential for water accumulation; however, its field-scale validation supported by high-resolution multispectral drone imagery remains limited. In agricultural systems, recurrent waterlogging can reduce productivity by impairing germination, prom... |
47. Optimization of Flight Parameters for High-Throughput Phenotyping of Guineagrass Using UASA fenotipagem de alto rendimento utilizando sistemas aéreos não tripulados (UAS) tornou-se uma ferramenta estratégica na agricultura de precisão aplicada a pastagens, permitindo a coleta rápida e não destrutiva de dados em larga escala. No entanto, a definição adequada dos parâmetros de voo, especialmente a distância de amostragem do solo (GSD) e a sobreposição de imagens, permanece um desafio, visto que configu... |
48. Integration of Spectral Phenological Markers and Artificial Neural Networks for Modeling the Yield of Potato CultivarsThe growing demand for food underscores the importance of essential crops such as potato. In this context, understanding yield dynamics is critical, and digital agriculture emerges as a key tool, enabling more efficient estimation of this variable without the need for destructive sampling. Accordingly, this study aimed to use orbital remote sensing combined with artificial intelligence algorithms to develop more accurate and precise models for potato yield prediction. Field data collection wa... S. Luns, J.B. Souza, B. , L. Conceicao da silva, R.P. Silva, V. Carreira |
49. Comparison of Real and Simulated GSD in the Estimation of Canopy Height in Maize Using RPAHigh-throughput phenotyping using remotely piloted aircraft (RPA) has become a strategic tool in precision agriculture, enabling rapid and non-destructive estimation of crop traits such as canopy height. Ground Sampling Distance (GSD) is a critical flight parameter in this process. Resolutions derived from smaller GSD values improve accuracy but increase flight time, image volume, and processing cost, whereas the opposite reduces these demands at a possible cost to accuracy. Simulating coarse... C. Ragalzi, M.J. Lima, L. Felipe, N. Guimarães, M.F. Santos |
50. Semi-Automatic Plot Segmentation for Crop Phenotyping Using Adaptive Spectral Indices and SAM3Yield trials and hill plots are widely used in plant breeding to evaluate large numbers of genotypes simultaneously. Extracting per-plot canopy boundaries from drone imagery is a key step in this process, but manual delineation is time-consuming, and rigid grid overlays do not account for true canopy boundaries. This paper presents an annotation-free pipeline for segmenting individual plots from multispectral drone imagery, requiring only approximate plot dimensions as input. The pipeline fir... |