Field Operations Research Articles

Long-distance hand tractor controlled by android system in Wetland Field

Modernization of agricultural practices is essential to overcome challenges such as decreasing agricultural labor and decreasing agricultural land in Indonesia. This research examines the development and performance of a hand tractor that is controlled remotely via an Android-based system using the NodeMCU ESP8266 module and a WiFi network. This study focuses on the technical and operational performance of hand tractors in wetlands. Modifications were made to the tractor, by installing three DC motors to control the clutch and gas levers. Tests were carried out on the control system accuracy and response time to assess the system's suitability for field operations. The research results show that the Android system has 100% accuracy in executing all commands given, but the response time decreases as the distance between the Android device and the WiFi module increases, with an average delay reaching 6,8 seconds for a maximum 100 m distance. Field efficiency testing shows that a tractor with Android control has a work efficiency of 70.48% in the plowing process and 75.33% in the harrowing process, lower than with manual control which reaches an efficiency of more than 95%. Despite response delays and reduced efficiency, the system is an important step toward agricultural mechanization, especially in labor-intensive wetlands.

Numerical Modeling of Tsunamis Generated by Subaerial, Partially Submerged, and Submarine Landslides

Using the existing two-dimensional experimental data and Open-source Fields Operation and Manipulation (OpenFOAM) software, this study performs a comprehensive comparative analysis of three types of landslide-generated tsunamis (subaerial, partially submerged, and submarine). The primary objective was to assess whether numerical simulations can accurately reproduce the experimental results of each type and to compare the predictive equations of the tsunami amplitudes derived from experimental and simulated data. The mesh size and dynamic viscosity parameters were initially optimized for a specific partially submerged landslide tsunami scenario and then applied across a broader range of experimental scenarios. Most of the simulated wave amplitudes remained within the 50% error margin, although significant discrepancies were observed between landslide types. When focusing on the crest amplitude of the first wave, the simulations of subaerial landslides least deviated from the experimental data, with a mean absolute percentage error of approximately 20%, versus approximately 40% for the partially submerged and submarine landslides. The predictive equations derived from the simulations closely matched those from the experimental data, confirming that OpenFOAM can effectively capture complex landslide–tsunami dynamics. Nonetheless, variations in the coefficients related to slope angles highlight the need for further calibration to enhance the simulation fidelity.

Algebraic algorithms for vector bundles over curves

In this paper, we represent vector bundles over a regular algebraic curve as pairs of lattices over the maximal orders of its function field and we give polynomial time algorithms for several tasks: computing determinants of vector bundles, kernels and images of global homomorphisms, isomorphisms between vector bundles, cohomology groups, extensions, and splitting into a direct sum of indecomposables. Most algorithms are deterministic except for computing isomorphisms when the base field is infinite. Some algorithms are only polynomial time if we may compute Hermite forms of pseudo-matrices in polynomial time. All algorithms rely exclusively on algebraic operations in function fields. For applications, we give an algorithm enumerating isomorphism classes of vector bundles on an elliptic curve, and to construct algebraic geometry codes over vector bundles. We implement all our algorithms into a SageMath package.

Foraging behaviour and pollinator dynamics in lucerne crop of the Bundelkhand region: Implications for crop management in the subtropics

Abstract Pollination stands out as the paramount ecosystem service facilitated by insect pollinators, providing benefits to over three quarters of the world’s major crops. The conservation and effective management of honeybees constitute integral components for ensuring the sustainable productivity of cross-pollinated crops. The present study aimed to document the potential pollinators of lucerne in the Bundelkhand region in India, which is known for its subtropical climatic conditions. Further, the foraging behaviour, including foraging rate and speed, was meticulously documented. A total of 13 insect species from seven families across three insect orders were observed visiting lucerne flowers. Hymenopterans comprised the majority of the floral visitors (73.48%), followed by Lepidoptera (26.51%) during the study period. Among them, Apis dorsata (Hymenoptera: Apidae) emerged as the most predominant floral visitor, exhibiting peak activity at 12:00 h with a density of 7.85 bees/m2/10 min. Surprisingly, among the major floral visitors, A. florea has the highest mean foraging rate (number of flowers visited/minute/forager), i.e., 11.56, 11.59 and 10.79 bees/flowers/minute during the initial flowering stage, peak flowering stage and late flowering stage, respectively. However, the foraging speed (time spent by the forager on each flower) was longest in A. dorsata, i.e., 5.05, 3.78 and 5.41 s/flower during initial, peak and late flowering stages, respectively. The findings emphasize the importance of enhancing honeybee activity to boost lucerne seed yield. Further, the outcomes of the study can be leveraged to optimize honeybee colony management, scheduling field operations such as pesticide application and weeding during periods of minimal honeybee activity.

Geological and Engineering Integration Fracturing Design and Optimization Study of Liushagang Formation in Weixinan Sag

The Weixinan Sag in the Beibuwan Basin is rich in shale oil resources. However, the reservoirs exhibit rapid phase changes, strong compartmentalization, thin individual layers, and high-frequency vertical variations in the thin interbedded sandstone and mudstone. These factors can restrict the height of hydraulic fracture propagation. Additionally, the low-porosity and low-permeability shale oil reservoirs face challenges such as low production rates and rapid decline. To address these issues, the Plannar3D full 3D fracturing model was used to simulate hydraulic fracture propagation and to study the main controlling factors for fracture propagation in the second member of the Liushagang Formation. Based on the concept of geological–engineering integration, a sweet spot evaluation was conducted to identify reservoirs with relatively better brittleness, reservoir properties, and oil content as the fracturing targets for horizontal wells. The UFM model was then applied to optimize fracturing parameters. This study indicates that the matrix-type oil shale has a high clay mineral content, resulting in a low Young’s modulus and poor brittleness. This makes hydraulic fracture propagation difficult and leads to less effective reservoir stimulation. In contrast, hydraulic fractures propagate more easily in high-brittleness interlayer-type oil shale. Therefore, it is recommended to prioritize the extraction of shale oil from interlayer-type oil shale reservoirs. The difference in interlayer stress is identified as the primary controlling factor for cross-layer fracture propagation in the study area. Based on the concept of geological–engineering integration, a sweet spot evaluation standard was established for the second member of the Liushagang Formation, considering both reservoir quality and engineering quality. Four sweet spot zones of interlayer-type oil shale reservoirs were identified according to this evaluation standard. To achieve uniform fracture initiation, a differentiated segment and cluster design was implemented for certain high-angle sections of well WZ11-6-5d. Interlayer-type oil shale was selected as the fracturing target, and the UFM was used for hydraulic fracture propagation simulation. Fracturing parameters were optimized with a focus on hydraulic fracture characteristics and the estimated ultimate recovery (EUR). The optimization results were as follows: a single-stage length of 50 m, cluster spacing of 15 m, pump injection rate of 10 m3/min, fluid intensity of 25 m3/m, and proppant intensity of 3.5 t/m. The application of these optimized fracturing parameters in field operations resulted in successful fracturing and the achievement of industrial oil flow.

Intelligent alarm system for river embankment seepage based on BILSTM

Currently, the alarm functions of existing levee seepage monitoring systems are limited to single-parameter monitoring and lack rate-of-change alarms and correlation alarms. This can lead to false alarms, missed alarms, equipment failures, or unnecessary downtime. To enhance the intelligence of levee safety monitoring and seepage alarms, a levee seepage intelligent alarm system based on a Bidirectional Long Short-Term Memory (BILSTM) network model was designed and implemented. Firstly, data cleaning and preprocessing are carried out on the engineering safety monitoring operation data to reduce the influence of dirty data such as outliers and repetitive values on the accuracy of alarms. Secondly, for the correlation between the piezometric tube water levels of the levee and the Yangtze River water levels, a correlation analysis based on Mutual Information (MI) theory was conducted to minimize the effect of piezometric tube water level change delays on correlation. Finally, the BILSTM model was used to predict trends in these potentially abnormal data intervals. Based on engineering application requirements, alarm thresholds were established, and a multi-level alarm module was developed. Field operation test results show that the proposed method can accurately predict the piezometric tube water levels of levees, achieving intelligent alarms within the engineering safety monitoring system.

Analysis on Innovation and Application of Anti-counterfeiting Technology of Printed Matter

This paper deeply focuses on the innovation and application of printed matter anti-counterfeiting technology, through a variety of advanced anti-counterfeiting technology principle, characteristics and rich application examples to carry out a detailed study, a comprehensive and in-depth analysis of its protection of product safety, resolutely prevent counterfeiting and fully maintain market order and other aspects of the key and important role. At the same time, the future development trend of anti-counterfeiting technology and the current severe challenges are deeply discussed, aiming at providing a very valuable and meaningful reference for the in-depth research and practical operation in related fields.

NeRSI: Neural Implicit Representations for 5D Seismic Data Interpolation

Due to challenging field operations and resource constraints, seismic data acquisition often requires coping with missing traces. Interpolation algorithms are crucial for reconstructing these missing traces to enable improved subsurface analysis and interpretation. While deep learning has made exciting advances in seismic reconstruction, its focus has predominantly been on 2D and 3D datasets with relatively low rates of missing data. Five-dimensional (5D) seismic data reconstruction entails considering simultaneous sources and receivers deployed in areal arrays to solve the reconstruction problem. The latter offers greater data redundancy, which can be leveraged to enhance interpolation quality. Traditional 5D deep learning interpolation methods rely heavily on synthetic training pairs, posing challenges when applied to real-world data. This necessitates transfer learning techniques, which can be cumbersome. Addressing this, we introduce a self-supervised, coordinate-based deep interpolation algorithm that obviates the need for labeled data. Employing a multi-layer perceptron (MLP) network can effectively encode the continuous seismic wavefield inherent in 5D data. Once trained, the MLP can infer missing trace amplitudes from their coordinates. We contribute to boosting the MLP, enabling it to generate seismic profiles rather than single-point predictions. This enhancement significantly strengthens the model’s performance and efficiency. Moreover, we apply nuclear norm regularization to the output profiles, improving the reconstruction quality. The effectiveness of our algorithm is supported by both synthetic and field data experiments, demonstrating its superior reconstruction capabilities.

Prototyping of vibration-sensing-actuation device which is contained high-power electrodynamic speaker for diagnosis of actual water main network

Water leakage and bursting due to deterioration of water distribution mains lead to water outages and economic losses. To prevent the accidents, we are conducting research to develop measurements and diagnostic methods for buried water mains. In this study, we focused on the shift in eigenfrequencies of in-plane bending vibration of rings because of deterioration. Therefore, we are developing the vibration-sensing-actuation device which detects the frequency changes of in-plane bending mode. The water distribution network is mainly composed from cast iron pipe. Moreover, the attached facilities for example fire hydrant and water valve are contained. Since attached facilities are fabricated from cast iron component, actual water main network is heavy facilities which have high rigidity. In order to excite the desired in-plane bending mode in heavy components, the proposed sensing-actuation device requires the high-power actuator. In this study, we introduce the high-power electromagnetic actuator more than 100W output into the proposed device. Furthermore, the fundamental operational tests were conducted under the various input signals in the laboratory situations. In addition, the field operation confirmation was performed in an actual water main network.

Design and Testing of an Electric Side-Mounted Cabbage Harvester

To address the limitations of current cabbage harvesters in China, which are often designed for a single variety and lack adaptability to different cabbage varieties, we developed an electric side-mounted cabbage harvester suitable for field operations in the Jiangsu region of China. This design is informed by the statistical analysis of the physical and agronomic parameters of major cabbage varieties. The harvester consists of key components, including an extraction device, a leaf-stripping device, a clamping and conveying device, and a root-cutting device. Powered by a 120 Ah direct current (DC) power source, it is capable of performing cabbage extraction, feeding, clamping, conveying, root cutting, and boxing in a single operation for three hours. Through theoretical analysis of the key components, specific parameters were determined, and field tests were conducted to verify the design. The results of the field experiments indicate that all components of the cabbage harvester operated effectively. Optimal performance was observed when the extraction roller speed was set between 100 and 110 RPM, the conveyor belt speed at 60 RPM, and the cutter speed between 160 and 220 RPM, resulting in a low cabbage harvest loss rate. The harvest loss rates from the three experiments were 11.3%, 13.3%, and 12%, respectively, which meets the mechanical harvesting requirements for cabbage.

Environmental benefits of ozonated water for sustainable grapevine disease control: A life cycle and carbon sequestration analysis

The oxidative capacity of ozone makes it a highly effective biocide, widely used in the food and processing industry, as well as in drinking water plants. In a context of tighter restrictions on the use of synthetic plant protection products at the regulatory level, wineries are looking for alternative methods to control diseases, making the application of ozonated water an option to consider. To determine the environmental sustainability of this alternative vineyard disease control treatment, an environmental impact assessment was conducted using the Life Cycle Assessment (LCA) methodology. The assessment was carried out in a vineyard located in the D.O. Rías Baixas of Galicia, located in northwestern Spain. The analysis was conducted on the basis of two functional units: i) 1 kg of grapes and ii) 1 ha of land managed and using a cradle-to-gate boundary system. Ten impact categories were assessed using the ReCiPe method, except for the water scarcity indicator, which was quantified using AWARE.The environmental sustainability of the vineyard was further analyzed by measuring its carbon sequestration potential to obtain a more complete environmental profile. The RothC-26.3 model was chosen to estimate absolute carbon sequestration and annual sequestration rate from 2020 to 2040. Since the ozone-only scenarios lost all harvest, no environmental assessment was performed for such scenarios. For the remaining scenarios, the findings suggest that those using a combination of ozonated water and fungicides have the worst environmental performance due to notable reductions in grape yield and more frequent disease control interventions, particularly in certain scenarios. When evaluating environmental performance per hectare of land managed, the scenario using ozonated water with a limited number of disease control interventions exhibited the most favorable environmental profile, primarily due to reduced use of fungicides and disease control passes. In addition, the main contributors to the vineyard environmental profile identified were diesel fuel combustion during field operations, fertilizer use and production, and irrigation. In addition, the research indicates that the vineyard sequesters carbon annually, but this alone is insufficient to offset its greenhouse gas emissions. However, it is estimated that through the application of more appropriate soil management practices, the vineyard could achieve carbon neutrality and potentially increase soil carbon stocks over time.

Research on the treatment and secondary fluid mixing technology for oil-containing drilling wastewater in gas fields

Drilling operations in oil and gas fields will generate a large amount of highly polluted and difficult to degrade drilling wastewater. Oil drilling wastewater mainly includes diluted drilling fluid mud, heavy metals and radioactive substances. At present, the treatment methods for oily drilling wastewater mainly include reinjection into the formation, recycling and discharge after standard treatment, but the treatment effect is not good. This article describes the treatment of oily drilling wastewater through wastewater oxidation, flocculation, mechanical dehydration, intermediate water oxidation, flocculation and filtration. The experimental results show that the RR of the wastewater can reach 93.50%, the maximum oil recovery rate (ORR) can reach 89.56%, the light transmittance (LT) of the purified water can reach 98.40%, the suspended solids (SS) can be reduced to 46.12mg/L, the HBO2- content can be reduced to 4.20mg/L. The water quality meets the requirements of DB61/T 1248-2019. Indoor technology was used for on-site testing, and the lowest reduction rate (RR) and ORR after wastewater treatment were 87.64% and 84.21%, respectively. Comparing the cost of treating the same type of wastewater, the treatment cost of the process in this article can be reduced by 1.87 USD/m3, which has higher economic benefits.

OPTIMIZATION SYSTEM FOR LOCAL CONTROL OF COMPRESSOR STATION

The article was devoted to the characteristics of using a compressor to process gas deposits. It is an important principle that the liquid is lifted to the surface of the earth by the energy of the gas compressed from the layer during processing. The structure of the well, the diameter of the exploitation belt, and the fountain pipes should be determined in such a way that there is little pressure in the well and in the trunk. Fountain pipes should not be lowered from the barrier to reducing pressure in the well's trunk and saving metal consumption. Depending on the daily gas volume of oil deposits, the gradient of pressure in the bed and the duration of the gas-water border should have been known. The change in those parameters depends primarily on the energy of the bed. The exploitation of the oil bed can lead to gas and flooding. The gas tank is the force that enables the liquid to move. In the compressor regime, however, the main energy is the energy of six gases. The compressor lifts the liquid to the surface of the earth, mainly because of air or gas. Keywords: gas compressor, field operation compress, compressor characteristics, formation water, wellbore pressure.

Medium-term economic impacts of cover crop adoption in Maryland

Cover cropping has the potential to generate private economic benefits to farm operations as well as larger-scale environmental benefits to the broader community. However, cover crop adoption remains limited in the United States (US) (i.e., 4.7 % in 2022 (USDA-NASS, 2024)), primarily due to uncertainty in economic outcomes, with several studies showing potentially negative net returns from cover crop use in the short-term (1–3 years). This study investigates the medium-term (5–7 years) economic impact of cover crop adoption using plot-level data from field experiments in the state of Maryland. The empirical analysis employs ordinary least squares (OLS) statistical models and partial budgeting techniques to achieve the study objective. Our results show that cover crops do not have a statistically significant effect on crop yield, fertilizer costs, or pesticide costs, but we find that cover crop use statistically increases field operation and seed expenses. As a result, the private net return from cover crop adoption in the medium term is generally negative based on the Maryland field trial data used in the analysis. Specifically, the average net return per acre with cover crops is approximately $60-$90 lower for corn and around $60 lower for soybeans compared to fields without cover crops. This empirical finding suggests that public support through incentive payments may help further incentivize cover crop adoption in the US, which can then provide environmental and ecosystem services that are of benefit to the general public.

Output speed control for hydro-mechanical continuously variable transmission of tractor.

In order to improve the speed stability of tractors equipped with hydro-mechanical continuously variable transmission (HMCVT) during field operations, the speed regulation characteristics, torque characteristics, and power diversion characteristics of HMCVT are analyzed. And a comprehensive control strategy for HMCVT output speed is proposed by analyzing the factors affecting the fluctuation of output speed, which combines engine speed control and displacement ratio control of hydraulic speed control system. This strategy adopts fuzzy control to regulate the engine speed to maintain the stability of engine operation. Moreover, adopting feed forward compensation control for speed compensation enhances the anti-interference ability of the hydraulic speed control system. And combined with model predictive control, adjust the swash plate swing angle of the variable pump to reduce the fluctuation of HMCVT output speed. By establishing joint simulation model, the effectiveness of the control strategy was verified using tractor acceleration mode and step load disturbance mode. The results show that the strategy can reduce the fluctuation rate of engine speed and impact degree, shorten the adjustment time, and improve the stability of tractor operating speed.

Development of a Machine Learning Framework for Real-Time PMI (Post-Mortem Interval) Estimation in Field Forensics

Accurate estimation of the Post-Mortem Interval (PMI) is critical in forensic investigations, aiding in determining the time of death. However, traditional PMI estimation methods, often reliant on physiological observations and environmental factors, face significant limitations in accuracy and efficiency, especially in field conditions. This paper presents the development of a machine learning (ML) framework designed for real-time PMI estimation, integrating multimodal sensor data to address the challenges encountered in field forensics. Our framework utilizes environmental and physiological features, including body temperature, ambient humidity, and biochemical decomposition markers, to predict PMI with high precision. The ML model, trained on historical forensic data, is deployed on a real-time processing platform, enabling rapid analysis and decision-making in resource- constrained environments. The system is optimized for field operations, incorporating low-power hardware and edge computing capabilities to provide forensic investigators with reliable PMI estimates on-site. Through a series of controlled experiments simulating forensic scenarios, our framework demonstrates a significant improvement in PMI accuracy compared to traditional methods, while maintaining low latency for real-time applications. This research highlights the potential of machine learning to revolutionize forensic practices, offering a scalable and adaptive solution for time-sensitive investigations. Here are some relevant keywords for the development of a machine learning framework for real-time PMI (Post- Mortem Interval) estimation in field forensics: Keywords: Field Forensics, Real-Time Machine Learning, Body Decomposition Stages, Machine Learning in Forensic Science, Artificial Intelligence for PMI Analysis, Sensor Data in PMI Estimation, Deep Learning for PMI Estimation, Automated Forensic Analysis, Data Acquisition in Field Forensics.

Kinematic analysis, workspace definition, and self-collision avoidance of a quasi-spherical parallel manipulator

Abstract Bilateral teleoperation has witnessed significant development since the mid-20th century, addressing challenges related to human presence in environments with constraints or a lack of skilled professionals. This article presents the kinematic and self-collision analyses of the quasi-spherical parallel manipulator, a three-legged parallel robot used as a haptic master device. The device is designed for remote center of motion-constrained operation in the telesurgical field. Inverse and forward kinematics are thoroughly analyzed to study working modes, singular configurations, and implement a haptic control architecture. The research explores the operative and reachable workspaces of the possible working modes, comparing them to find the most suitable one. Results highlight how the addition of the self-collision phenomenon impacts the working mode choice, drastically reducing most of the modes’ operative workspaces. An anti-collision control algorithm is finally introduced to maintain the architecture within its reachable workspace.

Rapid Detection of Soil Available Phosphorus using Capacitively Coupled Contactless Conductivity Detection

Background: In China, the traditional method for analyzing soil available phospho-rus is inadequate for large-scale soil assessment and nationwide soil formulation demands. To address this, we propose a rapid and reliable method for soil-available phosphorus detection. The setup includes an on-site rapid pre-treatment device, a non-contact conductivity detection device, and a capillary electrophoresis buffer solution system composed of glacial acetic acid and hydroxypropyl-β-cyclodextrin. Methods: The on-site rapid pre-treatment process includes fresh soil moisture content detec-tion (moisture rapid detector), weighing (handheld weighing meter), stirring (handheld rapid stirrer), and filtration (soil rapid filter) to obtain the liquid sample, and direct injection (capil-lary electrophoresis detector). The phosphate ion detection parameters include capillary size, separation voltage, injection parameters, and electric injection. We used Liaoning brown soil, Henan yellow tidal soil, Heilongjiang black soil, and Anhui tidal soil as standard samples. Additionally, we used mathematical modeling methods and machine learning algorithms to analyze and process research data. Results and Conclusion: Following calibration with standard samples, the experimental blind test samples demonstrated conformity with the national standard method, exhibiting a relative standard deviation of less than 3%. The proposed pre-treatment device and non-contact con-ductivity detector are powered by lithium-ion batteries, rendering them ideal for extended field operations. The non-contact conductivity detector obviates the need for direct contact with test samples, mitigating environmental pollution. Furthermore, the neural network model exhibited the highest level of goodness of fit in chemical data analysis.

Bud-YOLO: A Real-Time Accurate Detection Method of Cotton Top Buds in Cotton Fields

Cotton topping plays a crucial and indispensable role in controlling excessive growth and enhancing cotton production. This study aims to improve the operational efficiency and accuracy of cotton topping robots through a real-time and accurate cotton top bud detection algorithm tailored for field operation scenarios. We propose a lightweight structure based on YOLOv8n, replacing the C2f module with the Cross-Stage Partial Networks and Partial Convolution (CSPPC) module to minimize redundant computations and memory access. The network’s neck employs an Efficient Reparameterized Generalized-FPN (Efficient RepGFPN) to achieve high-precision detection without substantially increasing computational cost. Additionally, the loss calculation of the optimized prediction frame was addressed with the Inner CIoU loss function, thereby enhancing the precision of the model’s prediction box. Comparison experiments indicate that the Bud-YOLO model is highly effective for detecting cotton top buds, with an AP50 of 99.2%. This performance surpasses that of other YOLO variants, such as YOLOv5s and YOLOv10n, as well as the conventional Faster R-CNN model. Moreover, the Bud-YOLO model exhibits robust performance across various angles, occlusion conditions, and bud morphologies. This study offers technical insights to support the migration and deployment of the model on cotton topping machinery.

Driveway Detection for Weed Management in Cassava Plantation Fields in Thailand Using Ground Imagery Datasets and Deep Learning Models

Weeds reduce cassava root yields and infest furrow areas quickly. The use of mechanical weeders has been introduced in Thailand; however, manually aligning the weeders with each planting row and at headland turns is still challenging. It is critical to clear weeds on furrow slopes and driveways via mechanical weeders. Automation can support this difficult work for weed management via driveway detection. In this context, deep learning algorithms have the potential to train models to detect driveways through furrow image segmentation. Therefore, the purpose of this research was to develop an image segmentation model for automated weed control operations in cassava plantation fields. To achieve this, image datasets were obtained from various fields to aid weed detection models in automated weed management. Three models—Mask R-CNN, YOLACT, and YOLOv8n-seg—were used to construct the image segmentation model, and they were evaluated according to their precision, recall, and FPS. The results show that YOLOv8n-seg achieved the highest accuracy and FPS (114.94 FPS); however, it experienced issues with frame segmentation during video testing. In contrast, YOLACT had no segmentation issues in the video tests (23.45 FPS), indicating its potential for driveway segmentation in cassava plantations. In summary, image segmentation for detecting driveways can improve weed management in cassava fields, and the further automation of low-cost mechanical weeders in tropical climates can be performed based on the YOLACT algorithm.