Foundation models in smart agriculture: Basics, opportunities, and challenges

Computers and Electronics in Agriculture, Год журнала: 2024, Номер 222, С. 109032 - 109032

Опубликована: Май 29, 2024

Язык: Английский

---

WeedVision: A single-stage deep learning architecture to perform weed detection and segmentation using drone-acquired images

Computers and Electronics in Agriculture, Год журнала: 2024, Номер 219, С. 108792 - 108792

Опубликована: Март 2, 2024

Deep learning (DL) inspired models have achieved tremendous success in locating target weed species through bounding-box approach (single-stage models) or pixel-wise semantic segmentation (two-stage models), but not both. Therefore, the goal of this research study was to develop a single-stage DL architecture that only locate presence bounding-boxes also achieves instance on unmanned aerial system (UAS) acquired remote sensing images. Moreover, developed experiments integrating novel C3 and C3x module within its backbone for dense feature extraction, as well ProtoNet (Prototypical network) head component masking. Furthermore, proposed has been trained five categories dataset exported using multiple combinations various augmentation techniques, namely, C1, C2, C3, C4, C5, which metrics were assessed desktop graphical processing unit (GPU) palm-sized edge device (AGX Xavier). Results suggest category combination six data outperformed remaining by achieving precision scores 85.4 % (bounding-boxes) 82.8 (masking) GPU. Whereas, same model converted TorchScript able achieve 79.1 77 masking accuracy an device, respectively. The two potential applications when integrated with site-specific management technologies. First, it enables real-time detection, allowing immediate identification weeds spot-spraying applications. Second, facilitates masking, aiding estimation growth extent actual field conditions. combines both computer vision - detection – provide comprehensive information about growth, eliminating need algorithm.

Язык: Английский

---

CNN-MLP-Based Configurable Robotic Arm for Smart Agriculture

Agriculture, Год журнала: 2024, Номер 14(9), С. 1624 - 1624

Опубликована: Сен. 17, 2024

Amidst escalating global populations and dwindling arable lands, enhancing agricultural productivity sustainability is imperative. Addressing the inefficiencies of traditional agriculture, which struggles to meet demands large-scale production, this paper introduces a highly configurable smart robotic arm system (CARA), engineered using convolutional neural networks multilayer perceptron. CARA integrates arm, an image acquisition module, deep processing center, embodying convergence advanced robotics artificial intelligence facilitate precise efficient tasks including harvesting, pesticide application, crop inspection. Rigorous experimental validations confirm that significantly enhances operational efficiency, adapts seamlessly diverse contexts, bolsters precision farming practices. This study not only underscores vital role intelligent automation in modern agriculture but also sets precedent for future innovations.

Язык: Английский

---

Crop Detection and Maturity Classification Using a YOLOv5-Based Image Analysis

Emerging Science Journal, Год журнала: 2024, Номер 8(2), С. 496 - 512

Опубликована: Апрель 1, 2024

In recent years, the accurate identification of chili maturity stages has become essential for optimizing cultivation processes. Conventional methodologies, primarily reliant on manual assessments or rudimentary detection systems, often fall short reflecting plant’s natural environment, leading to inefficiencies and prolonged harvest periods. Such methods may be imprecise time-consuming. With rise computer vision pattern recognition technologies, new opportunities in image have emerged, offering solutions these challenges. This research proposes an affordable solution object classification, specifically through version 5 You Only Look Once (YOLOv5) model, determine location state rocoto peppers cultivated Ecuador. To enhance model’s efficacy, we introduce a novel dataset comprising images their authentic states, spanning both immature mature stages, all while preserving settings potential environmental impediments. methodology ensures that closely replicates real-world conditions encountered by system. Upon testing model with this dataset, it achieved accuracy 99.99% classification task 84% rate crops. These promising outcomes highlight potential, indicating game-changing technique small-scale farmers, especially Ecuador, prospects broader applications agriculture. Doi: 10.28991/ESJ-2024-08-02-08 Full Text: PDF

Язык: Английский

---

Pasture monitoring using remote sensing and machine learning: A review of methods and applications

Remote Sensing Applications Society and Environment, Год журнала: 2025, Номер unknown, С. 101459 - 101459

Опубликована: Янв. 1, 2025

Язык: Английский

---

Integrating reinforcement learning and large language models for crop production process management optimization and control through a new knowledge-based deep learning paradigm

Computers and Electronics in Agriculture, Год журнала: 2025, Номер 232, С. 110028 - 110028

Опубликована: Фев. 12, 2025

Язык: Английский

---

Fields of the Future: Digital Transformation in Smart Agriculture with Large Language Models and Generative AI

Computer Standards & Interfaces, Год журнала: 2025, Номер unknown, С. 104005 - 104005

Опубликована: Март 1, 2025

Язык: Английский

---

SoybeanNet: Transformer-based convolutional neural network for soybean pod counting from Unmanned Aerial Vehicle (UAV) images

Computers and Electronics in Agriculture, Год журнала: 2024, Номер 220, С. 108861 - 108861

Опубликована: Апрель 4, 2024

Язык: Английский

---

Performance evaluation of semi-supervised learning frameworks for multi-class weed detection

Frontiers in Plant Science, Год журнала: 2024, Номер 15

Опубликована: Авг. 20, 2024

Precision weed management (PWM), driven by machine vision and deep learning (DL) advancements, not only enhances agricultural product quality optimizes crop yield but also provides a sustainable alternative to herbicide use. However, existing DL-based algorithms on detection are mainly developed based supervised approaches, typically demanding large-scale datasets with manual-labeled annotations, which can be time-consuming labor-intensive. As such, label-efficient methods, especially semi-supervised learning, have gained increased attention in the broader domain of computer demonstrated promising performance. These methods aim utilize small number labeled data samples along great unlabeled develop high-performing models comparable counterpart trained large amount samples. In this study, we assess effectiveness framework for multi-class detection, employing two well-known object frameworks, namely FCOS (Fully Convolutional One-Stage Object Detection) Faster-RCNN (Faster Region-based Networks). Specifically, evaluate generalized student-teacher an improved pseudo-label generation module produce reliable pseudo-labels data. To enhance generalization, ensemble student network is employed facilitate training process. Experimental results show that proposed approach able achieve approximately 76% 96% accuracy as 10% CottonWeedDet3 CottonWeedDet12, respectively. We offer access source code (https://github.com/JiajiaLi04/SemiWeeds), contributing valuable resource ongoing research beyond.

Язык: Английский

---

Perspectives on the application of remote sensing technology in the cultivation of medicinal plants

International Journal of Remote Sensing, Год журнала: 2025, Номер unknown, С. 1 - 34

Опубликована: Янв. 16, 2025

Cultivation of medicinal plants (CMPs) plays a crucial role in sustaining the production resources (MPs). In light depletion wild plant (MPRs), CMPs have become primary source for meeting market demand. However, traditional methods are often limited, subjective, and time-sensitive. recent years, remote sensing (RS) has emerged as an important tool obtaining information on MPs, addressing many limitations inherent conventional techniques. This paper first highlights challenges faced provides comprehensive review main applications RS field. Subsequently, it summarizes existing analysing data, organizing findings previous studies according to types tasks methodologies employed. Approaches data analysis that could be applied Traditional Chinese Medicine (TCM) planning generalized compared. Finally, discusses potential difficulties cultivation process outlines future prospects technologies. latest research application can serve valuable resource both researchers practitioners. Additionally, offers curated selection those interested leveraging technologies precision agriculture plants.

Язык: Английский

---