Topics
| Filter results38 paper(s) found. |
|---|
1. AI-based Precision Weed Detection and EliminationWeeds are a significant challenge in agriculture, competing with crops for resources and reducing yields. Addressing this issue requires efficient and sustainable weed elimination systems. This paper presents a comprehensive overview of recent advancements in weed elimination system development, focusing on innovative technologies and methodologies. Specifically, it details the development and integration of a weed detection and elimination system based on the CoreXY architecture, implemented... H. Kulhandjian, M. Kulhandjian, D. Rocha, B. Bennett |
2. AgDataBox-IA – Web Application with Artificial Intelligence for Agricultural Data Analysis in Precision AgricultureAgriculture has been continually evolving, incorporating hardware, software, sensors, aerial surveys, soil sampling for chemical, physical, and granulometric analysis (based on sample grids), and microclimatic data, leading to a substantial volume of data. This requires platforms to store, manage, and transform these data into actionable information for decision-making in the field. In this regard, Artificial Intelligence (AI) is the most widely used tool globally to mine and transform vast d... R. Sobjak, C.L. Bazzi, K. Schenatto, W.K. Oliveira, A.E. Menegasso |
3. Multi-sensor Remote Sensing: an AI-driven Framework for Predicting Sugarcane FeedstockPredicting saccharine and bioenergy feedstocks in sugarcane enables stakeholders to determine the precise time and location for harvesting a better product in the field. Consequently, it can streamline workflows while enhancing the cost-effectiveness of full-scale production. On one hand, Brix, Purity, and total reducing sugars (TRS) can provide meaningful and reliable indicators of high-quality raw materials for industrial food and fuel processing. On the other hand, Cellulose, Hemicell... M. Barbosa, D. Duron, F. Rontani, G. Bortolon, B. Moreira, L. Oliveira, T. Setiyono, L. Shiratsuchi, R.P. Silva, K.H. Holland |
4. Drought Tolerance Assessment with Statistical and Deep Learning Models on Hyperspectral Images for High-throughput Plant PhenotypingDrought is an important factor that severely restricts blueberry growth, output and adversely impacts the desirable physiologic quality. Considering the challenges posed by climate change and erratic weather patterns, evaluating the drought tolerance of blueberry plants is not only vital for the agricultural industry but also for ensuring a consistent supply of these nutritious berries to consumers. Blueberry plants have a relatively ineffective water regulation mechanism due to their shallow... M. Rahman, S. Busby, A. Sanz-saez, S. Ru, T. Rehman |
5. A High-throughput Phenotyping System Evaluating Salt Stress Tolerance in Kale Plants Cultivated in Aquaponics EnvironmentsMonitoring plant growth in a controlled environment is crucial to make informed decisions for various management practices such as fertilization, weed control, and harvesting. Agronomic, physiological, and architectural traits in kale plants (Brassica oleracea) are important to producers, breeders, and researchers for assessing the performance of the plants under biotic and abiotic stresses. Traditionally, architectural, and morphological traits have been used to monitor plant growth. H... T. Rehman, M. Rahman, E. Ayipio, D. Lukwesa, J. Zheng, D. Wells, H.H. Syed |
6. Algorithm to Estimate Sorghum Grain Number from Panicles Using Images Collected with a Smartphone at Field-scaleAn estimation of on-farm yield before harvest is important to assist farmers on deciding additional input use, time to harvest, and options for end uses of the harvestable product. However, obtaining a rapid assessment of on-farm yield can be challenging, even more for sorghum (Sorghum bicolor L.) crop due to the complexity for accounting for the grain number at field-scale. One alternative to reduce labor is to develop a rapid assessment method employing computer vision and artificial intell... G.N. Nocera santiago, P. Cisdeli magalhães, I. Ciampitti, L. Marziotte |
7. Towards a Digital Peanut Profile Board: a Deep Learning ApproachArtificial intelligence techniques, particularly deep learning, offer promising avenues for revolutionizing object detection and counting algorithms in the context of digital agriculture. The challenges faced by peanut farmers, particularly the precise determination of optimal maturity for digging, have prompted innovative solutions. Traditionally, peanut maturity assessment has relied on the Peanut Maturity Index (PMI), employing a manual classification process with the aid of a peanut profi... M.F. Freire de oliveira, B.V. Ortiz, J.B. Souza, Y. Bao, E. Hanyabui |
8. Utilizing Hyperspectral Field Imagery for Accurate Southern Leaf Blight Severity Grading in CornCrop disease detection using traditional scouting and visual inspection approaches can be laborious and time-consuming. Timely detection of disease and its severity over large spatial regions is critical for minimizing significant yield losses. Hyperspectral imagery has been demonstrated as a useful tool for a broad assessment of crop health. The use of spectral bands from hyperspectral data to predict disease severity and progression has been shown to have the capability of enhancing e... G. Vincent, M. Kudenov, P. Balint-kurti, R. Dean, C.M. Williams |
9. HOPSY: Harvesting Optimization for Production of Strawberry Using Real-time Detection with YOLOv8Optimizing the harvesting process presents a continuous challenge within the strawberry industry, especially during peak seasons when precise labor allocation becomes critical for efficiency and cost-effectiveness. The conventional method for addressing this issue has been hindered by an absence of real-time data regarding yield distribution, resulting in less-than-ideal worker assignments and unnecessary expenditures on labor. In response, a novel, portable, real-time strawberry detection sy... Z. Huang, W. Lee, N. Takkellapati |
10. Optimal Placement of Soil Moisture Sensors in an Irrigated Corn FieldPrecision agricultural practices rely on characterization of spatially and temporally variable soil and crop properties to precisely synchronize inputs (water, fertilizer, etc.) to crop needs; thereby enhancing input use efficiency and farm profitability. Generally, the spatial dependency range for soil water content is shorter near the soil surface compared to deeper depths, suggesting a need for more sampling locations to accurately characterize near-surface soil water content. However, det... D. Mandal, L. Longchamps, R. Khosla |
11. Real-time Seed Mapping Using Direct MethodsSeed distance estimations are critical for planter evaluation and the prediction of planting parameter performance. However, these estimations are typically not conducted in real-time. In this study, we propose a real-time seed mapping approach using cameras and computer vision networks, augmented by a Kalman filter for vehicle state estimation. This process involves the transformation of pixel coordinates into real-world coordinates. We conduct a comparative analysis between these estimates ... A. Sharda, R. Harsha chepally |
12. Incorporating Return on Investment for Profit-driven Management ZonesAdopting site-specific management practices such as profitability zones can help to stabilize long-term profit while also favoring the environment. Profitability maps are used to standardize data by converting variables into economic values ($/ha) for different cropping systems within a field. Thus, profitability maps can be used to define management zones from several years of data and show the regions within a field which are more profitable to invest in for production, or those that can be... A.A. Boatswain jacques, A.B. Diallo, A. Cambouris, E. Lord, E. Fallon |
13. Leveraging UAV-based Hyperspectral Data and Machine Learning Techniques for the Detection of Powderly Mildew in VineyardsThis paper presents the development and validation of machine learning models for the detection of powdery mildew in vineyards. The models are trained and validated using custom datasets obtained from unmanned aerial vehicles (UAVs) equipped with a hyperspectral sensor that can collect images in visible/near-infrared (VNIR) and shortwave infrared (SWIR) wavelengths. The dataset consists of the images of vineyards with marked regions for powdery mildew, meticulously annotated using LabelImg.&n... S. Bhandari, M. Acosta, C. Cordova gonzalez, A. Raheja, A. Sherafat |
14. Combining Remote Sensing and Machine Learning to Estimate Peanut Photosynthetic ParametersThe environmental conditions in which plants are situated lead to changes in their photosynthetic rate. This alteration can be visualized by pigments (Chlorophyll and Carotenoids), causing changes in plant reflectance. The goal of this study was to evaluate the performance of different Machine Learning (ML) algorithms in estimating fluorescence and foliar pigments in irrigated and rainfed peanut production fields. The experiment was conducted in the southeast of Georgia in the United States i... C. Rossi, S.L. Almeida, M.N. Sysskind, L.A. Moreno, A. Felipe dos santos, L. Lacerda, G. Vellidis, C. Pilcon, T. Orlando costa barboza |
15. Using Machine Vision to Build Field Maps of Forage Quality and the Need for Agriculture-specific Machine Vision NetworksMachine vision systems have truly come of age over the past decade. These networks are relatively simple to implement with systems such as YOLOv5 or the more recent YOLOv8. They are also relatively easy and computationally cheap to retrain to a custom data set, allowing for customization of these networks to new object detection and classification tasks. With this ease, it is no surprise that we are seeing an explosion of these networks and their application through all aspects of a... P. Nugent, J. Neupane |
16. Detection of Sorghum Aphids with Advanced Machine VisionSorghum aphid, Melanaphis sorghi (Theobald), became a significant pest concern due to the significant yield losses caused in the sorghum production region. Different management practices, including monitoring and applying insecticides, have been used to manage this invasive pest in sorghum. The most common management strategy consists of visual assessments of aphids on sorghum leaves to determine an economic threshold level to spray. However, because of their rapid reproduction,... I.A. Grijalva teran, B. Spiesman, N. Clark, B. Mccornack |
17. Sparse Coding for Classification Via a Locality Regularizer: with Applications to AgricultureHigh-dimensional data is commonly encountered in various applications, including genomics, as well as image and video processing. Analyzing, computing, and visualizing such data pose significant challenges. Feature extraction methods become crucial in addressing these challenges by obtaining compressed representations that are suitable for analysis and downstream tasks. One effective technique along these lines is sparse coding, which involves representing data as a sparse linear combination ... A. Tasissa, L. Li, J.M. Murphy |
18. Spatial Predictive Modeling to Quantify Soybean Seed Quality Using Remote Sensing and Machine LearningIn recent years, the advancement of artificial intelligence technologies combined with satellite technology is revolutionized agriculture through the development of algorithms that help producers become more sustainable. This could improve the conditions of farmers not only by maximizing their production and minimizing environmental impact but also due to better economic benefits by allowing them to access high-value-added markets. Furthermore, the use of predictive tools that could improve t... C. Hernandez, P. Kyveryga, A. Correndo, A. Prestholt, I. Ciampitti |
19. Obstacle-aware UAV Flight Planning for Agricultural ApplicationsThe use of unmanned aerial vehicles (UAVs) has emerged as one of the most important transformational tools in modern agriculture, offering unprecedented opportunities for crop monitoring, management, and optimization. To ensure effective and safe navigation in agricultural environments, robust obstacle avoidance capabilities are required to mitigate collision risks and to ensure efficient operations. Mission planners for UAVs are typically responsible for verifying that the vehicle is followi... K. Joseph, S. Pitla, V. Muvva |
20. Supervised Hyperspectral Band Selection Using Texture Features for Classification of Citrus Leaf Diseases with YOLOv8Citrus greening disease (HLB), a disease caused by bacteria of the Candidatus Liberibacter group, is characterized by blotchy leaves and smaller fruits. Causing both premature fruit drop and eventual tree death, HLB is a novel and significant threat to the Florida citrus industry. Citrus canker is another serious disease caused by the bacterium Xanthomonas citri subsp. citri (syn. X. axonopodis pv. citri) and causes economic losses for growers from fruit drops and blemishes. Citrus cank... Q. Frederick, T. Burks, P.K. Yadav, M. Dewdney, J. Qin, M. Kim |
21. A Growth Stage Centric Approach to Field Scale Corn Yield Estimation by Leveraging Machine Learning Methods from Multimodal DataField scale yield estimation is labor-intensive, typically limited to a few samples in a given field, and often happens too late to inform any in-season agronomic treatments. In this study, we used meteorological data including growing degree days (GDD), photosynthetic active radiation (PAR), and rolling average of rainfall combined with hybrid relative maturity, organic matter, and weekly growth stage information from three small-plot research locati... L. Waltz, S. Katari, S. Khanal, T. Dill, C. Porter, O. Ortez, L. Lindsey, A. Nandi |
22. RMAPs: an Integrated Tool to Delimitate Homogeneous Management ZonesManagement zones are one of the most studied methods in precision agriculture to optimize crop yield from the soil, plant, management, and climate input parameters. We present Rmaps, an R package that integrates soil and crop yield spatial variability using geostatistical methods and one-hidden-layer perceptron (OHLP) to identify how input parameters influence crop yield and delimitate homogenous zones. From georeferenced data of soil, plant, management, climate, and crop yield parameters, Rm... E. Erazo, C. Mosquera, O. Ochoa |
23. AI Enabled Targeted Robotic Weed ManagementIn contemporary agriculture, effective weed management presents a considerable challenge necessitating innovative solutions. Traditional weed control methods often rely on the indiscriminate application of broad-spectrum herbicides, giving rise to environmental concerns and unintended crop damage. Our research addresses this challenge by introducing an innovative AI-enabled robotic system designed to identify and selectively target weeds in real-time. Utilizing the advanced Machine Learning t... A. Balabantaray, S. Pitla |
24. AI-enabled 3D Vision System for Rapid and Accurate Tree Trunk Detection and Diameter EstimationHuanglongbing (HLB) is the major threat to citrus production in Florida. Imidacloprid and oxytetracycline injections were proven to be effective in controlling HLB. The total amount of imidacloprid and oxytetracycline needs to be injected for the tree depending on the trunk diameter. Therefore, precisely measuring trunk diameter is important to effectively control the HLB. However, manually injecting imidacloprid or oxytetracycline and measuring the trunk diameter is time-consuming and labor-... C. Zhou, Y. Ampatzidis |
25. Active Learning-based Measurements Prediction in Sparsely Observed Agricultural FieldsThe sustainability of farming methods relies on the quality of soil health. Rich soil supplies vital nutrients to plants. The soil structure and aggregation possess crucial physical attributes that facilitate the infiltration of water and air, as well as enable roots to explore. Long-term and extensive monitoring of soil data is crucial for obtaining important information into the water dynamics of the land surface. Soil moisture dynamics play a critical role in the hydrothermal process that ... D. Agarwal, A. Tharzeen, B. Natarajan |
26. Cyberinfrastructure for Machine Learning Applications in Agriculture: Experiences, Analysis, and VisionAdvancements in machine learning algorithms and GPU computational speeds over the last decade have led to remarkable progress in the capabilities of machine learning. This progress has been so much that, in many domains, including agriculture, access to sufficiently diverse and high-quality datasets has become a limiting factor. While many agricultural use cases appear feasible with current compute resources and machine learning algorithms, the lack of software infrastructure for collec... L. Waltz, S. Khanal, S. Katari, C. Hong, A. Anup, J. Colbert, A. Potlapally, T. Dill, C. Porter, J. Engle, C. Stewart, H. Subramoni, R. Machiraju, O. Ortez, L. Lindsey, A. Nandi |
27. In-Field and Loading Crop: A Machine Learning Approach to Classify Machine Harvesting Operating ModeThis paper addresses the complex issue of classifying mode of operation (active, idle, stationary unloading, on-the-go unloading, turning) and coordinating agricultural machinery. Agricultural machinery operators must operate within a limited time window to optimize operational efficiency and reduce costs. Existing algorithms for classifying machinery operating modes often rely on heuristic methods. Examples include rules conditioned on machine speed, bearing angle and operational t... D. Buckmaster, J. Krogmeier, J. Evans, Y. Zhang, M. Glavin, D. Byrne, S.J. Harkin |
28. Estimating Real-time Soil Water Content (SWC) in Corn and Soybean Fields Using Machine Learning Models, Proximal Remote Sensing, and Weather DataSoil Water Content (SWC) is crucial for precise irrigation management, especially in center-pivot systems. Real-time estimation of SWC is vital for scheduling irrigation to prevent overwatering or underwatering. Proper irrigation yields benefits such as improved water efficiency, enhanced crop yield and quality, minimized environmental impact, optimized labor and energy costs, and improved soil health. Various in-situ techniques, such as Time-domain reflectometry (TDR), frequency-do... N. Chamara, Y. Ge, F. Bai |
29. Enhancing Seeding Efficiency: Evaluating Row Cleaners with Computer Vision in Precision AgricultureIn precision agriculture, the effective sowing of seeds is crucial but often hindered by challenges like hair pinning, low soil temperatures, and heavy residue on the soil surface. To address these issues, row cleaners are employed to clear the path for seeder opener discs, ensuring a clean, uniform trench for seed placement. This study examines the performance of various row cleaner models and introduces a novel method for their automatic, quantitative evaluation using computer vision techno... F. Sidharth, A. Sharda, B.G. Berretta |
30. Cotton Yield Estimation Using High-resolution Satellite Imagery Obtained from Planet SkySatSatellite images have been used to monitor and estimate crop yield. Over the years, significant improvements on spatial resolution have been made where ortho images can be generated at 30-centimeter resolution. In this study, we wanted to explore the potential use of Planet SKYSAT satellite system for cotton yield predictions. This system provided imagery data at 50 centimeters resolution, and we collected data 14 times during the season. The data were collected from two different cotton... M. Bhandari |
31. AIR-N: AI-Enabled Robotic Precision Nitrogen Management PlatformThe AI-Enabled Robotic Nitrogen Management (AIR-N) system is a versatile, cloud-based platform designed for precision nitrogen management in agriculture, targeting the reduction of nitrous oxide emissions as emphasized by the EPA. This end-to-end integrated system is adaptable to various cloud services, enhancing its applicability across different farming environments. AIR-N's framework consists of three primary components: a sensing layer for gathering data, a cloud layer where AI and ma... A. Kalra, S. Pitla, J.D. Luck |
32. Bird Welfare and Comfort in Poultry Coops Through Computations and AIBird welfare and comfort is very important inside poultry coops during transportation, especially during summer and winter months. The microenvironment inside a poultry coop resulting from hot/cold temperatures, relative humidity and heat production leads to complex scenarios affecting the bird welfare. The enthalpy comfort index (ECI) that relates to temperature, relative humidity was calculated to evaluate the poultry coop welfare that corresponds to bird welfare conditions (comfort; ... R. Pidaparti, A. Moghadham, H. Thippareddi |
33. Determining Desirable Swine Traits that Correlate to High Carcass Grades for Artificial Intelligence PredictionsWith the global population continuing to grow, there has been an increased stress applied to the agriculture industry to improve efficiency and yield. To achieve this goal within the cattle industry, selection and reproductive decisions have been lucrative aspects, both genetically and fiscally. Breeding animal selection impacts farms through passing on favorable market, reproductive, and temperament traits. The cattle industry has experienced genetic advancement due to the flexibility of art... A.N. Spina, J.P. Fulton, S.A. Shearer, T. Berger-wolf, D. Drewry |
34. Optimizing Soybean Management with UAV RGB and Multispectral Imagery: a Neural Network Method and Image ProcessingPrecision agriculture (PA) has emerged as a fundamental approach in contemporary agricultural management, aimed at maximizing efficiency in the use of resources and improving crop productivity. The transition to so-called "agriculture 4.0" represents a revolution in the way technology is applied in the field, with an emphasis on digital and automated solutions such as UAVs (Unmanned Aerial Vehicles). These devices offer new capabilities for capturing high-resolution images, enabling... F. Pereira de souza, L. Shiratsuchi, H. Tao, M. Acconcia dias, M. Barbosa, T. deri setiyono, S. Campos |
35. Premier Strategy Consulting - Sponsor Presentation... C. Zhu |
36. Advanced Classification of Beetle Doppelgängers Using Siamese Neural Networks and Imaging TechniquesThe precise identification of beetle species, especially those that have similar macrostructure and physical characteristics, is a challenging task in the field of entomology. The term "Beetle Doppelgängers" refers to species that exhibit almost indistinguishable macrostructural characteristics, which can complicate tasks in ecological studies, conservation efforts, and pest management. The core issue resides in their striking similarity, frequently confusing both experts and a... P.R. Armstrong, L.O. Pordesimo, K. Siliveru, A.R. Gerken, R.O. Serfa juan |
37. Spectral Imaging Deep Learning Mapper for Precision AgricultureWith the growing variety of RGB cameras, spectral sensors, and platforms like field robots or unmanned aerial vehicles (UAV) in precision agriculture, there is a demand for straightforward utilization of collected field data. In recent years, deep learning has gained significant attention and delivered impressive results in the realm of computer vision tasks, such as semantic segmentation. These models have also found extensive applications in research related to precision agriculture and spe... L. Thomas, B. Jakimow, A. Janz, P. Hostert, A. Lajunen |
38. Predicting Soybean Yield Using Remote Sensing and a Machine Learning ModelSoybean (Glycine max L.), a nutrient-rich legume crop, is an important resource for both livestock feed and human dietary needs. Accurate preharvest yield prediction of soybeans can help optimize harvesting strategies, enhance profitability, and improve sustainability. Soybean yield estimation is inherently complex because yield is influenced by many factors including growth patterns, varying crop physiological traits, soil properties, within-field variability, and weather conditions. The obj... M. Gardezi, O. Walsh, D. Joshi, S. Kumari, D.E. Clay, J. Rathore |