Search

• Reviewed "precision agriculture", "smart agriculture" and "digital agriculture" concepts. • Suggested "smart agriculture" and "digital agriculture" be phases of "precision agriculture". • Proposed a few research focuses on smart and digital technologies for agriculture. [ABSTRACT FROM AUTHOR]

• Proposed a corn seed spacing detection method based on image stitching and YOLOX. • The image stitching method based on "speed-rate" linkage could stitch multiple seed bed images into seed bed panoramic images. • YOLOX has been applied to the recognition and localization of seed targets in seed bed panoramic images. • This method can accurately detect the spacing of corn seeds in the seedbed panorama. Corn seed spacing is fundamental for detecting doubles and misses and calculating the seed spacing conformity and variation. These factors are essential for optimizing the corn seeding operation system and improving the quality of corn seeding. This paper proposed a corn seed spacing detection method based on image stitching and YOLOX in the seedbed. The image stitching algorithm based on the "speed-rate" linkage could generate a seedbed panorama by stitching the images obtained from an industrial camera. As a result, the seedbed panorama could present the overall distribution of seeds in the detection region. The YOLOX-based seed spacing detection algorithm was utilized to detect the seeds in the seedbed panorama, determine the position coordinates of each seed, as well as compute the seed spacing. This paper established a corn seed spacing detection system based on the above method and constructed the soil trough and field experiments. The soil trough experimental results indicated that: (1) The seedbed panorama generated by the image stitching algorithm based on the "speed-rate" linkage could accurately reflect the seedbed information; compared to other image stitching algorithms based on feature extraction, the image stitching algorithm based on the "speed-rate" linkage had obvious advantages in stitching speed and completeness of seed spacing information. (2) With a score threshold of 0.5, YOLOX had an accuracy of 96.67 %, a recall of 97.89 %, and an F1 of 97 %, which could be adapted for seed detection in seedbed images. (3) The missed detection rate of the YOLOX-based seed spacing detection algorithm in the seedbed panorama was 3.8 %. Compared to the seed spacing of the seedbed panorama was measured employing ImageJ (IM), the average error of the YOLOX-based seed spacing detection algorithm was 1.99 %. Comparing the seed spacing of the seedbed panorama obtained by the YOLOX-SSDA (YD) with IM, the average distinction between the two detection results was 1.53 mm, and the average detection error for seed spacing was 2.79 %. The field experimental results indicated that at different forward speeds (6 km/h, 8 km/h, and 10 km/h), corn seed spacing detection accuracy was 95.5 %, 93.2 %, and 90.6 %, respectively. Furthermore, the feasibility and accuracy of the corn seed spacing detection method in the seedbed were demonstrated. This research positively impacted the advancement of seeding detection technology and the enhancement of corn seeding quality. [ABSTRACT FROM AUTHOR]

• Proposed FSDnet has good performance in crop plant phenotype segmentation. • Introduced Gaussian density can reduce the calculation and improve accuracy. • By embedding FSDnet, the traditional models can work well in different densities. • We construction a new field 3D deep learning dataset including strawberry and apple. The accurate segmentation of fruit phenotypes in the field is of great significance for agricultural automation in the 3D scene. Although the existing fruit segmentation based on 3D point cloud has made great progress, in the complex field environment, due to lighting, leaf occlusion, shooting angle and other problems, the point cloud obtained by depth camera often has the problem of multiple voids and discrete points, which seriously affects the accurate segmentation of fruit phenotype. This paper proposes a embedding subnetwork FSDnet based on density-based feature extraction and feature propagation and embeds it in the novel segmentation networks, which effectively improves the segmentation accuracy of the point cloud phenotype in multi-hole and multi-discrete fruits, including (1) The density-based point cloud feature extraction and feature propagation theory is proposed to alleviate the problem of perception degradation in fruit edge point caused by discrete points and holes caused by imcomplete point cloud in the agriculture scene. (2) A density-adaptive embedding semantic segmentation framework FSDnet is proposed, and embedding the classical point cloud neural network can significantly improve the segmentation accuracy of the fruit phenotypes with multiple holes and discrete points in the traditional network. (3) This paper made a strawberry dataset and tested the designed new neural network on both strawberry and apple filed dataset. After FSDnet is embedded on different novel net, almost all net have been improved. We verified the performance of FSDnet in different density states in agricultural scenarios, mitigated the negative impact of density on segmentation accuracy, proving that it can adapt to different point cloud density in agricultural scenarios in comparison between Gaussian density and other two traditional density schemes, Gaussian density reduces the computational traffic (0.58G) of the network while maintaining similar performance to the other two densities, proving the superiority of assuming a Gaussian density. [ABSTRACT FROM AUTHOR]

• A bibliometric and content analysis was conducted to comprehensively review and analyze deep learning research in cropland field identification. • The paper discusses the challenges of deep learning-based research on cropland field identification, providing readers with insights and directions for future research in the field • This study fills the gap by providing a systematically summarized review of this research area. The cropland field (CF) is the basic unit of agricultural production and a key element of precision agriculture. High-precision delineations of CF boundaries provide a reliable data foundation for field labor and mechanized operations. In recent years, with the dual advancements in remote sensing satellite technology and artificial intelligence, enabling the extraction of CF information on a wide scale and with high precision, research on CF identification based on deep learning (DL) has emerged as a highly esteemed direction in this field. To comprehend the developmental trends within this field, this study employs bibliometric and content analysis methods to comprehensively review and analyze DL research in the field of CF identification from various perspectives. Initially, 93 relevant literature pieces were retrieved and screened from two databases, the Web of Science Core Collection and the Chinese Science Citation Database, for review. The previous studies underwent quantitative analysis using bibliometric software across five dimensions: publication year, literature type and publication journal, country, author, and keyword. Subsequently, we analyze the current status and trends of employing DL in the field of CF identification from four perspectives: remote sensing data sources, DL models, types of CF extraction results, and sample datasets. Simultaneously, we combed through current publicly available sample datasets and data products that can be referenced to produce sample datasets for CFs. Finally, the challenges and future research focus of DL-based CF identification research are discussed. This paper provides both qualitative and quantitative analyses of research on DL-based CF identification, elucidating the current status, development trends, challenges, and future research focuses. [ABSTRACT FROM AUTHOR]

Current technology has made it possible to automate a number of agricultural processes that were traditionally carried out by humans and now can be entirely performed by robotic platforms. However, there are certain tasks like soft fruit harvesting, where human skills are still required. In this case, the robot's job is to cooperate/collaborate with human workers to alleviate their physical workload and improve harvesting efficiency. To accomplish that in a safe and reliable way, the robot should incorporate a safety system whose main goal is to reduce the risk of harming human co-workers during close human–robot interaction (HRI). In this context, this paper presents a theoretical study, addressing the safety risks of using collaborative robots in agricultural scenarios, especially in HRI situations when the robot's safety system is not completely reliable and a component may fail. The agricultural scenarios discussed in this paper include automatic harvesting, logistics operations, crop monitoring, and plant treatment using UV-C light. A human injury assessment is conducted based on converting the HRI in each agricultural scenario into a formal mathematical representation. This representation is later implemented in a probabilistic model-checking tool. We then use this tool to perform a sensitivity analysis that allows us to determine the probability that a human may get injured according to the occurrence of failures in the robot's safety or perception systems. Results of the sensitivity analysis show that an agricultural robot with a robust human perception system can still harm people if they are not well-trained to interact with the robot for certain scenarios. This illustrates how the probabilistic modeling methodology presented in this work can be used by safety engineers as a guideline to construct their own HRI models and then use the results of the model-checking to enhance the safety and reliability of their robot's safety system architectures and on-site safety policies. • Templates for hazard analysis and mitigation strategies in cooperative agricultural robotics. • Human–robot interactions in agricultural scenarios can be represented as simplified probabilistic models. • The reliability of robot safety systems/policies can be tested by formal verification before real implementation. • An agricultural robot with a robust human perception system can still harm people if they are not well-trained. • Ensuring that only trained people interact with agricultural robots, minimizes chances of producing human injuries. [ABSTRACT FROM AUTHOR]

Precision agriculture has emerged as a dominant force in the United States, with widespread adoption of advanced technologies and decision support systems (DSS) since the 1980s. Key tools such as variable rate application (VRA), autopilot systems, and remote sensing have become integral for U.S. farmers, offering invaluable insights from crop, soil, and weather information to optimize agricultural production while minimizing environmental impact. To synthesize and categorize the extensive research available on precision agriculture, a systematic review protocol has been designed. Our objective is to offer clear and authoritative insights into the nature, scope, and volume of this field. Implementing a rigorous search strategy, we utilized renowned databases such as Scopus® and Web of ScienceTM to gather relevant and significant materiality. The retrieval process involved the use of indexing terms and Boolean operators, with a focus on 'precision agriculture' and 'precision farming', striking a balance between specificity and comprehensiveness. To ensure the credibility of our findings, only peer-reviewed papers authored by individuals affiliated with U.S. institutions have been included. Expert reviewers with deep knowledge in the field independently assessed the selected papers, thoroughly evaluating titles, abstracts, keywords, methods, conclusions, and declarations. Consistency and eligibility were paramount in determining which papers met the criteria for inclusion. Any discrepancies or disagreements were resolved through rigorous consensus-building discussions among the reviewers. Through this comprehensive meta-review, we provide a scientific contribution that enhances our understanding of precision agriculture, highlighting focus areas for further research and development (R&D). By synthesizing and categorizing the existing literature, we offer authoritative insights into the research landscape, informing future investigations and fostering innovation. Focusing specifically on the U.S., we shed light on the unique aspects and pioneering advancements in precision agriculture within the country. Ultimately, our findings have the potential to drive progress, contributing to sustainable development, increased productivity, enhanced environmental sustainability, and responsible agricultural practices. [ABSTRACT FROM AUTHOR]

In current plant phenotyping research, the study of plant time-series images based on deep learning has received widespread attention. While such image data is relatively easy to obtain, the cost of annotation is high. One efficient method for achieving cost-effective training is through contrastive learning. Plant growth is slow, and the changes in image sequences over a period of time are small, with simple semantic information. Previous contrastive pre-training models struggled to effectively distinguish positive samples from the same image with different augmented views and similar negative samples from different images. Therefore, this paper proposes a method called self-supervised contrastive learning method for plant time-series images with a Priori Distance Embedding (PDE). The semantic information in images corresponding to different phenological stages of plants varies. This method transforms this crucial domain knowledge into prior distances for image pairs and conducts contrastive learning pre-training. The learned weights can be transferred to downstream tasks. Building upon this method, experiments were conducted on cherry time-series images to assess the quality of pre-training through a plant phenotyping image semantic segmentation task. To provide a comprehensive example of plant time-series image phenotypic analysis, this paper establishes a cherry growth temporal model, specifically including PDE pre-training, anomaly detection, semantic segmentation, and recording the results from the temporal dimension. The experiments indicate that this self-supervised contrastive learning method can be effectively applied to the pre-training of plant time-series images, demonstrating broad applicability in various computer vision studies related to plant phenotyping. • A pre-training method for plant images with a Priori Distance Embedding (PDE). • Transforms the phenological information into prior distances for image pairs. • An image acquisition device was built to obtain cherry time series images. • The PDE method was tested by plant phenotype image segmentation task. • We provided an exhaustive example of cherry time-series image phenotypic analysis. [ABSTRACT FROM AUTHOR]

• An advanced measurement method enables precision farming. • A precise, stable and novel live chickens shank length measurement method. • Fused images provide richer features. • An improved ResNet model applied to a regression task. Phenotypic parameters are crucial reference indicators in poultry breeding. However, the chickens shank length is still manually measured, which is time-consuming and labor-intensive. Additionally, the measurement results are difficult to unify due to the subjective factors of different individuals. To address this issue, this paper proposed a method for live chicken shank length measurement (SLM). It enriches chicken shank feature by fusing visible images and infrared images. The fusion images are then input into a deep regression model based on the improved ResNet. The measurement model used ResNet as its backbone and introduces Squeeze-and-Excitation (SE) blocks and a Spatial Pyramid Pooling (SPP) block, resulting in more precise and stable shank length measurements. The average coefficient of variation, average floating error, average standard deviation and Pearson correlation coefficient for shank length measurements using the fusion images are 0.21 %, 0.49 %, 0.181 mm and 0.996, respectively, compared with using single visible or infrared image, the accuracy and stability are obviously improved. That indicated combining deep learning model and fusion information, the SLM proposed in this paper can achieve a more precise, reliable and standardized measurement of live chicken shank length. [ABSTRACT FROM AUTHOR]

In facility horticulture smart farms, harvesting robotic systems to automate harvesting tasks are challenging due to the complex environment, irregular growth and fruits pose. To harvest a fruits, the vision information required by a harvesting robot is accurate position and pose. Especially, fruit pose information is essential for planning paths to avoid damaging stems, leaves, branches, and obstacles, preventing damage to the fruit and harvesting robots, and planning efficient harvest sequences. This paper presents a method that uses information from tomato sepals to estimate the orientation of tomatoes, which are a commonly grown crop in horticulture. First, we train a YOLOv8 model to detect and segment bodies and sepals of tomatoes. For robust training of the model, the training data is constructed using effective data augmentation, which synthesizes segmented foreground objects and inserting them into the background. Then, for finding association between the body and sepal of a tomato, we apply IoU score based matching and the Hungarian algorithm. Consequently, we obtain point clouds for both parts using RGB-D data. Finally, we compute the center point of each object by using spherical fitting and the statistics of the point cloud, respectively. Then, we estimate the pose of the tomato as a vector between two center points of the body and sepal. To accurately and practically evaluate the proposed method, we generated ground truth data using calibration patterns in tomato greenhouse. As the experimental results show, The segmentation results show that A P 50 , s e p a l is 94.7, A P 50 , t o m a t o is 96.3, and m A P is 61.5. The result of the pose estimation for the pose validation dataset is a mean error angle of 6.79 ± 3.18 and angle errors of less than 10 degrees account for 87.2%. The total algorithm proposed in this paper requires about 0.038s for each tomato greenhouse image. As a result, our proposed pose estimation can be practically utilized in robot systems for tomato harvesting. • Utilizing the sepal information for tomato pose estimation. • Hungarian algorithm and IoU to find association of tomato body and sepal. • Pose estimation as a direct vector from the center points of tomato body and sepal. • New ground truth using geometric computation for accurate pose verification. [ABSTRACT FROM AUTHOR]

• First seriously consider the signal filtering problem in broiler health monitoring research. • Newly propose a wavelet denoising algorithm for broiler sound signal filtering. • Newly propose the continuous differentiable threshold function and scale threshold. • Newly propose a two-stage variable multi-threshold pulse extraction algorithm. • Newly propose to evaluate signal filtering methods using classification accuracy. There is little attention paid to signal filtering in existing broiler health monitoring or broiler sound signal classification research, and the only few studies still have issues such as lack of specific, in-depth, and specialized details. In response to these problems, the authors conducted a depth analysis of the signal characteristics of effective signal components and noises, and proposed a broiler sound signal filtering method based on improved wavelet denoising and effective pulse extraction. This method consists of two parts. The first part is wavelet denoising, which is used to remove the widely distributed noises in broiler sound signals. Specifically, a wavelet denoising algorithm based on continuous differentiable threshold function and scale threshold has been proposed. The disadvantages in existing hard threshold function, soft threshold function, and semi-soft threshold function have been effectively addressed, and the smoothness of broiler sound signals has been ensured. The second part is pulse extraction processing, which is used to remove sudden noises that appears as independent pulses in broiler sound signals. Specifically, a two-stage variable multi-threshold pulse extraction algorithm has been proposed. The main advantages of existing three-threshold pulse extraction algorithm and double-threshold pulse extraction algorithm have been referenced, the threshold setting standard has been clarified, the continuous multi-pulse problem has been given special attention, and the effective signal components in broiler sound signals have been accurately extracted. In this way, the noises in broiler sound signals were removed, and the effective signal components were highlighted. The signal-to-noise ratio (SNR) and root mean square error (RMSE) from signal detection field, as well as the classification accuracy and F1-score from machine learning field, were used to comprehensively evaluate the effect of the signal filtering method proposed in this paper. A large number of tests shown that, compared to existing signal filtering methods, the broiler sound signal processed by the signal filtering method proposed in this paper achieved the maximum SNR and the minimum RMSE. On feature data sets created from broiler sound signals processed by different signal filtering methods, three classifiers all achieved the highest classification accuracy and F1-score on the feature data set corresponding to the signal filtering method proposed in this paper. The feasibility, practicality, and superiority of the signal filtering method proposed in this paper have been widely verified. This study is an important supplement to the broiler health monitoring and broiler sound signal classification research, providing important references for signal filtering research in similar fields. [ABSTRACT FROM AUTHOR]

• This paper introduced the functional role of robotic arms in precision agriculture. • This paper summarized the hardware and software technologies of agricultural robotic arms. • This paper presented the application of robotic arms in various agricultural environments, including greenhouses, fields, and orchards. • This paper discussed the challenges and future prospects of robotic arms in precision agriculture. In precision agriculture, robotic arms exhibit significant technical advantages, such as enhancing operational precision and efficiency, reducing labor costs, and supporting environmental sustainability. This paper provides a comprehensive overview of the application of ground-based robotic arms in precision agriculture, analyzing the hardware and software aspects and current application status across various agricultural settings, and discussing challenges and prospects in this field. First, this paper explores precision agriculture and agricultural robotic arms, highlighting their critical roles in enhancing agricultural efficiency and automation. Further, it addresses the challenges plaguing the practical applications of robotic arms and compares innovative robotic arm technologies with traditional models to establish a foundation for understanding these advancements in modern agriculture. Additionally, this paper analyses the hardware of robotic arms, including rigid and flexible manipulators, drivers, end-effectors, sensors, and controllers, emphasizing the importance of innovation and optimization for improved performance. For the software systems, this paper focused on classic workflows and advanced algorithms for perception, motion planning, and control, as these are essential for the precise and adaptable functioning of robotic arms in diverse agricultural environments. Furthermore, this paper reviews the research and application status of robotic arms across various settings, including greenhouses (e.g., ground planting, desktop planting, and vertical planting), fields (e.g., dry fields, moist, and paddy fields), and orchards (e.g., fruit tree orchards, vineyard orchards, and ground-level orchards) to demonstrate their broad applicability and efficient operational capabilities in diverse conditions. Lastly, this paper discusses the challenges and prospects of robotic arms, emphasizing the significance of integrating disciplines, such as agronomy and biomimetics, big data, artificial intelligence, digital twinning, and human–machine interaction. Advancements in these areas are pivotal for the progress of robotic arm technology and for introducing innovative, efficient solutions to precision agriculture. In summary, this review reveals the immense potential of the application of robotic arms in precision agriculture. With ongoing technological advancements, these robotic arms are expected to play an increasingly crucial role in future agricultural production, making substantial contributions to achieving more efficient, innovative, and sustainable farming practices, heralding a new era in agricultural technology. This paper will serve as a valuable guide for researchers and practitioners, offering comprehensive insights into the use of robotic arms in precision agriculture and providing essential knowledge for advancing the field. [ABSTRACT FROM AUTHOR]

• We proposed a delineating field parcel model (TSANet) based on satellite image time series. • TSANet learn the relevance of spatial-spectral-temporal feature representation. • TSANet is robust across space and time. • TSANet performs better than pervious methods. • The time series data from the main growing period has a greater impact. Satellite image time series (SITS), such as Sentinel-2 imagery, plays a crucial role in the delineation of agricultural fields by reducing the impacts of ambiguities due to the spatial arrangement of field boundaries. Existing delineate field parcel models rely extensively on spatial features derived from single-date imagery. However, several studies have exploited the potential of SITS to effectively tackle the complexities associated with the intrinsic consistency between agricultural fields and their boundaries. This paper proposes a novel Two-Stream Attention convolutional neural network (TSANet) to capture the subtle difference between agricultural fields from SITS. Specifically, a field temporal semantic stream is introduced to adaptively leverage the significance of spatial-spectral-temporal feature representation associated with the location of agricultural parcels, especially where transitions in crop types take place. Considering the consistency between field parcels and their boundaries, we developed a field boundary prediction stream to enhance the extraction of edge features, particularly for the extraction of small and irregular agricultural parcels. Moreover, a field parcel refining block is employed to further enhance the geometric accuracy of agricultural fields. We conducted experiments on Sentinel-2 images from the Netherlands. Results showed that our approach produced a better layout of agricultural fields, with an average F1-score of 0.91 than the existing 3D-UNet, U-TEA, and BiConvLSTM. In addition, through the analysis of both quantitative and qualitative results, the stronger robustness of the model compared to other algorithms has been verified by temporal transfer and large-scale spatial prediction. We compared the difference between SITS and the corresponding composite images, which further verified the influence of temporal variation on the proposed approach. This paper provides a general guide for delineating agricultural parcels using SITS. [ABSTRACT FROM AUTHOR]

• Based on the control input cost function of the traditional partial format model-free adaptive control, an effective greenhouse temperature controller is designed by adding an index term that limits energy consumption. • The multi-parameter sensitivity analysis method based on Monte Carlo is used to analyze the sensitivity of the controller parameters to determine which parameters have a greater impact on system performance. Then, the whale optimization algorithm is used to optimize the sensitive controller parameters. This improves the efficiency of parameter tuning. • The improved partial format model-free adaptive controller achieves a balance between control accuracy and low energy consumption. Compared with the traditional partial format model-free adaptive controller, the energy consumption of the improved partial format model-free adaptive controller is reduced by 12.35 %. Temperature is critical to the growth of crops in agricultural greenhouses. Thus, designing a greenhouse temperature controller that maximizes energy savings while maintaining control accuracy is very important. This paper proposes an improved partial format model-free adaptive control method and designs a greenhouse temperature controller based on this method to balance control accuracy and energy consumption. Firstly, a limited energy consumption term is added to the control input cost function of the traditional partial format model-free adaptive control to penalize excessive control input. We derive an improved partial format model-free adaptive control input algorithm and design a greenhouse temperature controller using this algorithm. Then, the controller's sensitive parameters are selected using a Monte Carlo-based parameter sensitivity method. Finally, the whale optimization algorithm is used to optimize the sensitive parameters. The insensitive parameters are set according to experience. This paper, a theoretical study based on simulated experiments, proves the convergent tracking error of the proposed improved partial format model-free adaptive algorithm. Simulation results show that the improved controller minimizes energy consumption while ensuring control accuracy, reducing power consumption by 12.35 % compared to the traditional controller. [ABSTRACT FROM AUTHOR]

Soybean is a critical source of food, protein, and oil, and thus has received extensive research aimed at enhancing their yield, refining cultivation practices, and advancing soybean breeding techniques. Within this context, soybean pod counting plays an essential role in understanding and optimizing production. Despite recent advancements, the development of a robust pod-counting algorithm capable of performing effectively in real-field conditions remains a significant challenge. This paper presents a pioneering work of accurate soybean pod counting utilizing unmanned aerial vehicle (UAV) images captured from actual soybean fields in Michigan, USA. Specifically, this paper presents SoybeanNet, a novel point-based counting network that harnesses powerful transformer backbones for simultaneous soybean pod counting and localization with high accuracy. In addition, a new dataset of UAV-acquired images for soybean pod counting was created and open-sourced, consisting of 113 drone images with more than 260k manually annotated soybean pods. The images are taken from an altitude of approximately 13 ft, with angles between 53 and 58 degrees, under natural lighting conditions. Through comprehensive evaluations, SoybeanNet demonstrates superior performance over five state-of-the-art approaches when tested on the collected images. Remarkably, SoybeanNet achieves a counting accuracy of 84.51% when tested on the testing dataset, attesting to its efficacy in real-world scenarios. The publication also provides both the source code and the labeled soybean dataset, offering a valuable resource for future research endeavors in soybean pod counting and related fields. • Created and open-sourced the 1st UAV-acquired image dataset for soybean pod counting. • Pioneered the first study on precise soybean pod counting utilizing UAV images. • Demonstrated SoybeanNet's superiority over 5 state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

• A novel camera-integrated spraying needle nozzle design was developed. • Two pan-tilt units assembled together to provide 360-degree spraying capability. • The coordinates of all weeds were determined and spraying process was carried out. • Positional error tests were done to assess the targeting performance of the system. • Deposition experiments were done to evaluate the spraying efficiency of the system. While pesticide use is very important for weed control in agriculture, it is critical due to environmental contamination. In this study, a novel camera-integrated spraying needle nozzle design for targeted micro-dose spraying in precision weed control was developed in order to avoid excessive pesticide use. The micro-dose spraying system consisted of a camera, two pan-tilt units with servomotors assembled together to provide a 360-degree spraying capability for artificial weed samples, and spraying equipment. All the system automation and image processing processes were evaluated and controlled using LabVIEW software. The shooting capability and spraying performance of the novel camera-integrated spraying needle nozzle were tested and evaluated under laboratory conditions using artificial weed samples placed on a conveyor belt. The greenness method was used to detect the artificial weed targets on the conveyor belt. After image capturing, the coordinates of all artificial weeds in the field of view were calculated, and micro-dose spraying was then carried out for each artificial weed sample one by one until all the samples were sprayed. Positional error tests were carried out to assess the targeting performance of the spraying system. Deposition experiments were also carried out using filter papers to evaluate the spraying efficiency of the micro-dose spraying system under 200 kPa spraying pressure. A conveyor belt was set up for carrying the filter papers. The positional error test results showed that the weeding mean positional errors were 7.72 mm, 22.40 mm and 23.34 mm for the centre, left and right sides of the conveyor belt, respectively. The deposition concentration results showed that, while the mean deposition was 1.923 ng ml−1 for the centre, the mean depositions were 1.567 ng ml−1 and 1.494 ng ml−1 for left and right sides, respectively. Higher spraying efficiency was determined in the centre compared to the left and right sides according to the amount of deposition results. The novel micro-dose spraying system has been experimentally tested and found to be very efficient. [ABSTRACT FROM AUTHOR]

• Unmanned aerial spraying systems (UASS) with variable rate spraying (VRS) explored for reducing pesticide ecological impact. • Examines the integration of UASS in variable rate spraying as a new dimension in precision agriculture. • Highlighted the benefits of VRS in increasing agricultural yields, reducing resource usage, and enhancing environmental sustainability. • Multispectral and hyperspectral sensor technologies are investigated for optimizing UASS-based VRS missions. • Addresses technical and regulatory challenges in UASS-based spraying, offering potential solutions. • Future innovations in sensor technology and AI-driven flow rate optimization for precision agriculture are discussed. Pesticides suppress pest populations and maintain agricultural yield; however, their overuse causes ecological damage. To mitigate the ecological damage caused by the overuse of pesticides, this paper explores the application of a precise and adaptable technique known as unmanned aerial spraying system (UASS)-based variable rate spraying (VRS). Herein, the current state of precision agriculture is examined, and the application of UASS in variable rate spraying is discussed. Then, the role of advanced sensors, including multispectral and hyperspectral technologies, in optimizing UASS-based VRS missions is studied, followed by the challenges in UASS-based spraying, ranging from technical intricacies to regulatory considerations, and related solutions. In addition to delving into pesticides, the paper explores alternative solutions such as herbicides, encompassing an integrated approach that aligns with sustainable farming practices for more effective pest management. VRS has advantages, including increased yields, reduced resource usage, and environmental sustainability. This paper concludes by delineating current challenges and envisioning future innovations, spotlighting ongoing studies on sensor technology and AI-driven flow rate optimization. This review provides insights into UASS-based VRS for precision agriculture. [ABSTRACT FROM AUTHOR]

• The paper proposes a methodology for a hybrid model that can capture complex non-linear relationships and help understand the interactions between input and output features. • The model overcomes the limitation of traditional neural network models in predicting across different spaces. • The proposed model is an intelligent water quality prediction model that can provide decision support for predicting COD Mn concentration and managing watershed risks in the Changdang Lake basin, with good application prospects and practical value. The organic pollution of lake water can cause a tremendous threat to the water ecosystem and human health. The COD Mn is one of the crucial indicators of lake water quality and is commonly utilized to gauge the extent of organic pollution in lake water. Therefore, this paper selected COD Mn as the research object and used the water quality monitoring data of Changdang Lake in China and its upstream and downstream to predict the COD Mn concentration in the lake. In order to study the spatial relationship between the lake and upstream and downstream water quality, reflect the joint action of multiple water quality factors in prediction and the interaction between different feature factors. This study combined the XGBoost feature filtering algorithm, maximum mutual information coefficient (MIC), and improved recurrent neural network (GRU) and proposes a hybrid model called XGB-MIC-GRU. The model first used XGBoost to screen and extract the relative importance of water quality characteristics and used the Shapley addition extension (SHAP) method to explain XGBoost feature extraction. Then, the correlation between the lake and the upstream and downstream water quality is calculated through MIC analysis. Finally, the selected water quality factor characteristics and spatial characteristics are input into the GRU model for prediction. The experimental results showed that water temperature, total phosphorus, and total nitrogen are the most important to COD Mn , and the upstream US1 and downstream DS1 and DS2 stations are the most closely related to the concentration of COD Mn in the lake. By comparing the prediction effect of the model in different time steps, the best 16-time steps related data were selected to predict the value of the next time. MAE, RMSE, and R2 of the model are 0.10, 0.13, and 0.96, respectively. The model has better prediction accuracy and correlation error than the traditional SVR and GPR. The proposed mixed model can accurately predict the concentration of COD Mn in the lake. It can assist decision-makers in timely implementation of effective measures to safeguard the lake ecosystem. [ABSTRACT FROM AUTHOR]

Intelligent grazing is a new livestock husbandry development mode based on obtaining the multi-source information, ecological balance is the goal. This paper focuses on intelligent grazing, reviews from grass remote sensing and aerial seeding, unmanned aerial vehicles (UAVs) monitoring and intelligent grazing robot, and summarizing the development of intelligent grazing elements at this stage, exploring the new development direction of automatic grazing management with grazing robot at this stage, and proposes a novel intelligent grazing strategy based on remote sensing, herd perception and UAVs monitoring. Furthermore, focusing on the intelligent grazing needs, the digital twin pasture is built in this paper mainly serves the technical verification and data acquisition based on machine vision, simulating the behavior of the herd during grazing. • Reviewing from UAVs grazing, pasture remote sensing, livestock monitoring. • Proposing a grazing strategy with remote sensing, herd perception and UAVs monitoring. • Constructing a digital twin pasture to simulate the grazing herds. [ABSTRACT FROM AUTHOR]

• An efficient method is proposed for orchard tree trunk detection in natural scenes. • A multi-scenes orchard tree trunk dataset is established for experiments. • The YOLOv5 model structure is improved to enhance detection accuracy and speed. • Proposed method is robust against complex orchard environment. The recognition and localization of fruit tree trunks in orchard are important for orchard operation robots, which are the bases for automatic navigation, fruit tree spraying and fertilization etc. A method was proposed based on machine vision to detect target objects such as fruit tree trunks, person and supporters in orchard by improving the YOLOv5 deep learning algorithm in this paper, which is applicable to the recognition tasks in natural orchard environments. Firstly, 1354 images of the natural orchard collected by camera were image enhanced, weather effects such as rain, snow, bright light, shadow and fog were added to expand the dataset and to increase the robustness of the model. Secondly, the original YOLOv5 model was improved by replacing the Bottleneck network in the C3 module with the lightweight GhostNet V2 to reduce the network parameters, and changing the box loss function CIoU to SIoU in the loss function to make the regression of the detection box more accurate, and coordinate attention mechanism (CA) was added to the network to reduce the interference of useless background information in images. Before training, pre-anchor boxes were generated by using IoU-based K-means clustering, after that the dataset was fed into the improved YOLOv5 for training, and the trained model was used to detect the trunks. Finally, weighted boxes fusion (WBF) was used instead of the non-maximum suppression (NMS) in this paper for the output of the detection boxes. Then the density-based spatial clustering of applications with noise (DBSCAN) algorithm was used for trunk clustering. The improved target detection method was trained and validated on the experimental dataset. The model size is reduced by 43.6 %, model parameters are reduced by 46.9 %, and the mAP reaches 97.1 %, with an average detection speed of 198.2 ms per image. Compared with the original YOLOv5, the model is more lightweight, the detection accuracy and speed are improved. The improved YOLOv5 is also better than YOLOv3, NanoDet and SSD in terms of combined accuracy and speed, and has similar performance to YOLO_MobileNet in orchard dataset. The experimental results show that the improved YOLOv5 target detection model proposed in this paper is lightweight while still having better detection accuracy and detection speed in complex environments, and the model is small enough to be deployed to mobile or low-performance terminals for target detection in natural orchard environments. [ABSTRACT FROM AUTHOR]

In this work, we perform a systematic literature review of Open Data and Public Domain datasets in Agriculture. We use the PRISMA method to analyze the existing academic literature about open data in agriculture, concretely 1401 papers from the IEEE Xplore and Web of Science collections, published from 2012 to 2022. Many of these articles use or make available datasets of very different typologies, like sensor data, statistical data or satellite images, among others kinds. Some papers talk about different relevant topics that influence the use of open data, like the lack of open data for research purposes, barriers to adopting or sharing data, privacy concerns, data-sharing recommendations and guidelines. In addition, with the help of a script created ad-hoc for this research work, we analyze the degree of compliance within the FAIR principles of all the public domain or open datasets that we have been able to identify from the literature, concretely 104 datasets. The script can check the compliance of Gen2 maturity indicators of a list of resources. Using these metrics, we have been able to identify those open datasets that might be at risk of stopping to be available and started a "rescue operation". For those datasets whose terms permitted it, we migrated those Open Data datasets to Zenodo, a repository that complies with the availability principles. This will ensure the future survival of valuable data in Agriculture. • More than 70% of agriculture's public datasets come from satellite sources. • ESA members Canada and United States concentrate the most agriculture datasets. • Lack of research data in agriculture is a relevant issue in spite of initiatives to remove barriers and empower open data and public domain. [ABSTRACT FROM AUTHOR]

• Standardized ear tag data via transformation and cleaning. • Conducted time series analysis on daily pig data. • Analyzed seasonal trends in pig population data. • Classified ear tag data using six methods. The behavioral and physiological patterns of pigs are key indicators for assessing their health status. This study aims to explore the differences in activity level and body temperature variation patterns between normal and abnormal pigs. In view of the characteristics of pig ear tag data, such as high volatility, unclear features and strong randomness, we adopted various processing and analysis methods. The main work of this paper includes: (1) Preprocessing and integrating pig ear tag data to improve data quality; (2) Performing similarity and time series analysis on the daily data of each pig to identify the change patterns of activity level and body temperature; (3) Performing seasonal feature analysis and trend change analysis on the data of the pig population, finding the gap between abnormal pigs and the population pattern, and classifying abnormal and normal pigs. The main results and conclusions of this paper are as follows: (1) Pigs are most active in two time periods: 5:00 to 10:00 in the morning and 14:00 to 18:00 in the afternoon; (2) The activity duration of lame pigs is significantly lower than that of normal pigs, while the activity duration of pigs under other abnormal conditions is not much different; (3) Although the ear tag temperature data is affected by the pig's behavior, abnormal pigs can be found by the characteristic of low temperature trend within a day. (4) Random forest algorithm can effectively classify normal and abnormal pigs, with an accuracy of 0.879. The research results of this paper have important significance and value for realizing the refined management of pigs, improving the breeding efficiency, and ensuring the health and production safety of pigs. [ABSTRACT FROM AUTHOR]

• The purpose of this paper is to achieve the appearance quality inspection of citrus, addressing the challenges of defect target and fruit morphology detection. • A detection algorithm for defect targets on citrus peel and a method for fruit morphology detection have been proposed. • A citrus quality detection data set is constructed and evaluated. Experimental results show that the method in this paper achieves satisfactory results in performance. Identifying defects in citrus peels and analyzing fruit morphology are two core challenges in citrus quality inspection. In order to more accurately identify minor defects on citrus peels, we proposed a detection model Yolo-FD (Yolo for defects). The model was based on the Yolov5 network framework, and the backbone network embedded the Three-dimensional Coordinate Attention (TDCA) mechanism innovatively designed in this study. It accurately captured the subtle changes and feature associations of the target in spatial location, significantly enhancing the model's ability to perceive defects in fruit peels. Moreover, we employed a simplified Bidirectional Weighted Feature Pyramid Network (BiFPN) in the model to achieve cross-scale connections and improve the feature fusion ability of the model. At the same time, Contextual Transformer block (COT) was introduced into Neck network and the CoT3 module was built to fully capture the static and dynamic contextual information in the citrus defects images and enhance the expression of the feature map. Through this series of improvement methods, missed detections and false detections caused by small targets were effectively reduced. Fruit morphology detection was combined with the Partice Swarm Optimized Extreme Learning Machine (PSO-ELM) model to determine whether the citrus fruit morphology was well-formed, using the symmetry index, roundness and tilt angle of the citrus as input parameters. The experimental results indicated that the mean average precision of the Yolo-FD model is 98.7 % (mAP-0.5). Compared with Yolov5s, Yolov7-tiny, and Yolov8n, the mAP was improved by 1.4 %, 1.5 %, and 0.5 % respectively. Its average detection time for a single frame image on the server was 19.5 ms. And the PSO-ELM model achieved a fruit morphology detection accuracy of 91.42 %, a coefficient of determination of 0.9044, and a mean squared error of 0.8497. The research results met the accuracy and real-time requirements for citrus sorting on the production line, and could provide an effective solution for citrus grading and quality assessment. [ABSTRACT FROM AUTHOR]

• This paper presents a Bayesian multi-task learning model for pest detection. • The model estimates size of aphids, which is an indicator of infestation severity. • It also quantifies the predictive uncertainties enabling confident decisions. • It performs thorough audit studies that offer insight for future model refinements. Aphids are persistent insect pests that severely impact agricultural productivity. The detection of aphid infestations is critical for mitigating their effects. This paper presents an artificial intelligence approach to detect aphids in crop images captured by consumer-grade RGB imaging cameras. In addition to detecting the presence of aphids, the size of the aphid is an important indicator of infestation severity. To address these, we present a Bayesian multi-task learning model to detect the presence of aphids and estimate their size simultaneously. Our model employs a joint loss function, combining a classification loss and a customised size loss. The classification component aims to identify images containing aphids, whilst the customised size loss function estimates the size of the aphids. The latter is specifically designed to account for discrepancies between the estimated and actual ground truth sizes, enhancing the accuracy of the size estimation. The model utilizes a ResNet18 backbone, ensuring robustness and adaptability across various conditions. The proposed model was evaluated using an agricultural pest dataset consisting of images of corn, rape, rice, and wheat crops. It achieved aphid presence detection accuracies of 75.77%, 66.39%, 70.01%, and 59% for corn, rape, rice, and wheat images, respectively. An in-depth evaluation of predictive uncertainties revealed areas of high confidence and potential inaccuracies for both size and presence of aphids in images, offering insight for future model refinement. We also conducted an ablation study to thoroughly analyse the contributions of each component in proposed model. Our model offers a valuable tool that can be used in pest management strategies for facilitating more sustainable and efficient agricultural practices. [ABSTRACT FROM AUTHOR]

With the development of modern animal husbandry, especially pig farming, large-scale, intensive, and automated farming has become the trend in the industry. The realization of accurate recognition and warning of dangerous actions in feeding scenarios for sows and piglets holds high research and practical value in this field. This paper addresses this issue by proposing a Transformer-based Neural Network (TNN) model. This model emphasizes the extraction of global features and handling of long-distance dependencies, significantly improving the accuracy of behavior recognition. Compared with traditional neural network models, TNN demonstrates superiority in dealing with animal behavior recognition in farming scenarios. Furthermore, an in-depth study of the attention mechanism in the TNN model was conducted in this paper. By visualizing the attention heat map of the TNN model, it was found that TNN could effectively focus on key areas in the image, thereby accurately identifying the behavior of piglets. Finally, this paper proposes a unique model lightweighting strategy that allows the TNN model to run efficiently on edge devices. In the experimental part, the performance of the TNN model on five behavior recognition tasks was evaluated first. The results showed that the TNN model achieved high scores in both Precision and Recall, far exceeding traditional neural network models. Then, by visualizing the attention heat map of the TNN model, the superiority of the TNN model in focusing on key image areas was further confirmed. In the end, the effect of the model lightweighting strategy was demonstrated, and even with a significant reduction in parameters and computational complexity, the performance of the TNN model remained excellent. This research not only promotes the development of animal behavior recognition technology but also provides new insights and tools for the precise management and efficient operation of the animal husbandry industry. • A comprehensive dataset of sow nursing scenarios was collected and annotated. • A variety of data augmentation methods were proposed to rebalance it, ultimately leading to a model accuracy of 0.91 mAP. • A time-series neural network (TNN) model is proposed, tailored to the characteristics of the actual farm collection equipment. • An alarm triggering mechanism was designed, enabling the model's application in actual agricultural scenarios. [ABSTRACT FROM AUTHOR]

• A novel Optical Droplet Edge Imaging method for droplet measurement is proposed. • A fast in-situ measuring portable device for droplet deposition is designed. • An image processing algorithm is proposed to obtain deposition performance indicator. • The droplet deposited area to volume is calibrated with average deviation 3.22%. • This device outperforms the water sensitive paper in terms of ease of use and digitization. Accurate and rapid measurement of droplet deposition on the spray target provides feedback data on spray quality to optimise application parameters and confirm crop protection efficacy. However, simultaneously meeting the requirements for field measurement remains a challenge. In this paper, a novel Optical Droplet Edge Imaging (ODEI) method based on a portable droplet deposition and distribution measurement device is proposed to enable in-situ reusable sampling and on-line automatic digitisation for field spraying. First, the principle of the ODEI method is presented to realise the sampling state closer to droplet deposited on the leaves. Then, a low-cost portable in-situ droplet deposition measurement device is designed and the optimal light field parameters of the device are determined. Finally, the simultaneous acquisition of the top and bottom side of the droplet deposition image of the collector is obtained and its image processing algorithm is presented to obtain the performance indexes of droplet deposition, such as individual droplet size and distribution, total droplet number, volume median diameter, droplet density and the ratio of droplet amount between the top and bottom side of the collector. The relationship between droplet volume and image area is calibrated with an average deviation of 3.22 %. Experimental results show that the VMD deviation between the device and the WSP image measured by same image processing method is 15.34 μm. Due to the difference in deposition state between the device and the water sensitive paper (WSP), the droplet deposition area conversion from this device to the WSP was also calibrated. The absolute deviation between the WSP moving average coverage rate and the converted collector coverage rate of this device is 1.21 %. The measurement result of the spray experiment shows that this device has the same measurement capability as the WSP, as the droplet diameter distribution is very consistent with the WSP distribution. This in-situ and automatic measurement method is a promising and useful addition to the spray performance monitoring tool for field droplet deposition. [ABSTRACT FROM AUTHOR]

• Calyx localization method proposed in this paper obtained the best localization. • YOLOv5x improved detection accuracy of feature points in binocular images. • Pairing of kiwifruit and calyx facilitated calyx matching for kiwifruit localization. • Calyx matching mechanism effectively avoided feature points mismatching in global matching method. • Template matching with parallel polar line constraint improved success rate of kiwifruit matching. Localization is the first critical step for picking robots to successfully grasp fruit. However, classical binocular localization methods adopted global matching for kiwifruit, which may result in a large amount of mismatching feature points in complex orchard and thus cause low localization accuracy. Therefore, an improved binocular localization method of calyxes based on deep learning was proposed to accurately detect and locate kiwifruit for robotic harvesting. Calyxes in the binocular images and kiwifruit in the left images of the binocular images were detected using You Only Look Once version 5 xlarge (YOLOv5x). The detected calyxes were matched in the binocular images using kiwifruit and calyx pairing and kiwifruit matching. The matched calyxes in the binocular images were used to locate calyxes using three localization methods. Specifically, three binocular localization methods, i.e., calyx localization (CL), fruit localization (FL), and depth information from depth map (DIDM), were compared to find the optimal one. Ground truth three-dimensional coordinates of calyxes was measured by laser rangefinder and coordinate paper on a self-designed experimental platform. Results showed that YOLOv5x achieved an average precision (AP) of 99.5 % on kiwifruit detection and a mean AP of 93.5 % on kiwifruit and calyx detection with a detection speed of 108 ms per image. Average deviation of X-axis, Y-axis, and Z-axis obtained by the CL method were 7.9 mm, 6.4 mm, and 4.8 mm, respectively. Compared with the FL and DIDM methods, localization error rate of the proposed CL method was reduced by 55.1 % and 53.8 %, respectively. It indicates that the proposed CL method is promising for robotic harvesting. [ABSTRACT FROM AUTHOR]

• This paper establishes an image dataset of sika deer obtained from different perspectives. • This paper proposes an AD-YOLOv5 model to detect key parts of the sika deer. • This paper improves the loss function of the model to enhance both the training and inference ability of the model. • This paper proposes an adaptive channel attention mechanism for better extracting shallow features of the key parts of sika deer. • By altering the feature fusion network and combining it with the channel attention mechanism, the feature information is balanced across different scales, which enhances the accuracy of detection. Accurate and rapid fixation of key parts of sika deer is essential for the mechanized and intelligent collection of both deer blood and velvet antler. To solve the low detection accuracy problem of key parts of deer caused by the diversity of deer farm breeding environment and the rapid movement of deer, this paper proposes an AD-YOLOv5 algorithm for detecting key parts of sika deer based on YOLOv5s. Specifically, a new weighted bidirectional feature pyramid network S-BiFPN is first proposed, which both simplifies the structure and effectively performs skip feature fusion to better capture object information at different scales, so that the adaptive feature fusion capability of the model is optimized. Second, this paper proposes to add a SENet module after the C3 module of the 9th Layer in the backbone part of the network, so that the network can adaptively readjust the channel weights of feature maps to enhance the important features for key parts of sika deer and improve the expression ability of the model for key parts. Furthermore, to make the model more stable in the training process, a novel bounding box regression loss function SIoU is led into, which can better learn the position and size of the bounding box by fusing the orientation information between the ground truth box and the predicted box. The effectiveness of the proposed algorithm is demonstrated through ablation experiments. Experimental results show that the mAP @0.5 of the proposed algorithm reaches 97.30 % for the detection of key parts of sika deer, which is 12.85 %, 9.65 %, 9.18 %, 7.13 %, 3.59 %, 4.6 %, 0.90 % and 0.10 % higher than SSD, EfficientDet, YOLOv3, YOLOv4, Faster R-CNN, YOLOv5s, YOLOv7 and YOLOv8s, respectively. This research provides a new idea for detecting key parts of animals in complex environments and lays a theoretical foundation for the development of intelligent fixation devices for sika deer. [ABSTRACT FROM AUTHOR]

[Display omitted] • An image-based system for monitoring pests in mango orchards was developed. • System overcame bottlenecks of pest monitoring in remote and outdoor environment. • Deep learning algorithm detects and recognizes mango pests with an F 1 -score of 0.96. • More than 2 years of spatiotemporal data reflected trends in pest behavior. • Spatiotemporal data was used to realize new data-driven IPM strategies. Mango production is a prominent tropical fruit industry worldwide. However, outdoor mango cultivation is susceptible to crop damage caused by insect pests and harsh environmental conditions. Integrated pest management (IPM) has emerged as a proposed solution to this problem. IPM utilizes data-driven and environmentally-friendly methods to suppress insect pest populations. Nevertheless, the collection of insect pest population data remains a laborious process, necessitating automation. This paper presents an image-based monitoring system to automatically record insect pest populations and environmental conditions in mango orchards. The system comprises solar-powered sensor nodes capable of periodically acquiring and analyzing sticky paper trap images. A modular deep learning-based algorithm was developed to detect and classify insect pests into seven classes, including major insect pests of mango such as thrips, mango leafhopper, and oriental fruit fly, with an average classification F 1 -score of 0.96. Unlike other insect counting algorithms, the algorithm reliably classifies insect pests according to different taxonomic levels even in non-laboratory environments. The monitoring system was tested and deployed in a remote mango orchard for over two years. The collected spatiotemporal information was analyzed to demonstrate the benefits of using the proposed system and recommend new IPM strategies. Temporal data analysis revealed a significant decrease in the count of selected insect pests after using the system, enabling identification of insect hotspots through statistical methods. This work presents a breakthrough in hardware and software solutions for developing smarter insect pest monitoring systems, leading to better IPM strategies. [ABSTRACT FROM AUTHOR]

• A comprehensive overview of crop row detection method in crop management and canopy phenotyping. • Focus on visual perception methods for seedling stage crop row in field ground operation scenarios. • Representative devices, models, and algorithms for the crop row detection process are analyzed. • Discussed challenges and possible solutions to crop row detection. Crop row detection technology is widely used in field management operations, crop phenotyping, and other fields. The crop row detection method obtains guidance information from field roads through relative positioning. This approach can effectively mitigate the constraints associated with absolute positioning methods, such as GNSS guidance, and ultimately improve the autonomous positioning ability and navigation accuracy of agricultural equipment. Researchers have extensively explored methods for quickly, accurately, and robustly identifying the orientation features of crop rows within field environments. As a result, many algorithms and models have been established based on vision technology. The primary aim of this paper is to provide a comprehensive overview of the status of crop row detection methods, with a specific emphasis on visual perception techniques for seedling crops utilized in field ground autonomous navigation operations. The typical process of the crop row detection method is summarized in detail, including crop image data acquisition, crop canopy feature extraction, and crop row centerline detection. The representative devices, models and algorithms involved in each process are analyzed. The challenge of crop row detection and the restrictive issues of current research methods are discussed. The future research direction and potential solutions are proposed. [ABSTRACT FROM AUTHOR]

Burning agricultural straw after reaping is a typical farming approach for quicker crop rotation but compromises the sustainability of cultivation practices. Unmanned aerial vehicle (UAV) remote sensing techniques are regarded as a feasible coping strategy to the sustainability dilemmas confronted in the domain. In this paper, we evaluate current multisensor information fusion technologies including both traditional methods and deep learning approaches for remote monitoring in real agricultural scenes and investigate their applicability for detecting burning rice straw. To this end, we collect StrawBurning , a real-world dataset containing fully paired infrared and visible images through mapping ground scenes based on UAV remote sensing. Furthermore, we propose a novel multiscale contrast adaptation (MCA) method for efficient multisensor image fusion and accurate straw burning detection in real farmland scenarios. The MCA remarkably enhanced the detection performance of YOLOv5 on the StrawBurning data by approximately 3%, 2%, and 5%, in terms of recall, mAP@0.5 and mAP@0.5:0.95, respectively. Therefore, our proposed method is demonstrated to obtain the superior performance compared to the single-modal information and other advanced multisensor information fusion methods. Experimental results indicate that neural network models utilizing the proposed multisensor fusion data showed the potential to accurately detect burning rice straw. • A real-world straw burning dataset is published for agricultural sustainability. • Efficient training-free image fusion algorithm based on multiscale contrast adaptation. • Bi-modal inputs outperform the single-modal one in straw burning detection. • Suitable for future UAV on-board processors in smart agriculture. [ABSTRACT FROM AUTHOR]

• Precision agriculture method applicable to remote on-tree pomegranate pixel-based fruit size monitoring. • Comparison of Deep Learning and zero-shot vision Transformer instance segmentation models. • Image-Based size estimation algorithm for occluded fruits with protuberances. In situ automatic fruit monitoring is of great interest for more accurate and cost-efficient decision making in agriculture. For this purpose, the development of computer vision-based tools is essential. Deep Learning techniques have shown good performance in fruit detection and segmentation. Recently, new models based on Transformer architecture have emerged with promising potential and zero-shot inference capability. In this paper, a Deep Learning model, Mask R-CNN, was trained for on-tree pomegranate fruit segmentation and compared with foundational models based on Vision Transformer, Grounding DINO and Segment Anything Model. Results with Mask R-CNN proved a better performance, according to F1 score and AP metrics, and a lower computational cost, according to prediction time. One of the most interesting derived applications from fruit segmentation is fruit size estimation. However, segmented fruit masks are frequently incomplete due to occlusions. Therefore, image fruit size estimation is not a straightforward process. In this work, we also propose a novel algorithm to estimate and monitor the fruit size in pixel units from the automated masks. A median relative error of 1.39% was obtained, demonstrating the potential and feasibility of future fully-automatic fruit size estimators. [ABSTRACT FROM AUTHOR]

• The first time to grade density of the hybrid rice machine-transplanted blanket-seedling before transplanting. • Reliability of improved extremely randomized trees of grading the seedling under various sowing rates. • Effectiveness of seedling density grading for improved field planting performances. Due to the influence of seed conditions and environmental factors on growing process of rice seedlings, there is a significant difference in the density of rice seedlings during planting, thereby, each row of multiple rows transplanter cannot take the accurate number of seedlings in transplanting and thus affect the performance of mechanized transplanting. Accurately grading the density of Hybrid Rice Machine-Transplanted Blanket-Seedlings (HRMTBS) before transplanting can improve mechanized transplanting performance by using seedlings of the same density level in the multiple rows of the transplanter. First, based on Unmanned Aerial Vehicle（UAV） multi-source data, the Vegetation Indices (VI) of spectral images and the Texture Features (TF) of RGB images were extracted in this paper. Then, an Improved Extremely Randomized Trees (IERT) algorithm, which introduces bootstrap sampling, random selection of feature subsets, and optimal node splitting, promoted the speed and achieved accurate grading of HRMTBS's density before mechanized transplanting. By selecting the optimal kernel for TF extraction, the effects of Support Vector Machine (SVM), eXtreme Gradient Boosting (XGBoost), Extremely Randomized Trees (ERT), and IERT algorithm on seedling density grading were compared and analyzed. HRMTBS's density grading was conducted under different seeding rates of 50 g/tray, 70 g/tray, and 90 g/tray, which was used to verify the generalization ability of the IERT algorithm. Furthermore, based on the maximum information coefficient method, VIs and TFs were optimized, which simplified the classification model and significantly improved classification speed. The experimental results showed that the optimal kernel for TF extraction is 3 × 3. The IERT algorithm is superior to SVM, XGBoost, and ERT under three different seeding rates, with overall density grading accuracy of 95.4 %, 92.3 %, and 90.1 %, respectively. Refining VIs from spectral images and TFs information from RGB images effectively improves the algorithm's accuracy. Based on the IERT algorithm, the overall grading accuracy of three different seeding rates increased by 9.4 %, 8.0 % and 7.6 %, and 6.5 %, 5.7 % and 5.1 % for classification using multi-source data compared to using only spectral images (VIs) or RGB images (TFs), respectively. Under the condition of ensuring classification accuracy, the optimized TFs and VIs reduced the number of input model features from 18 to 5 and optimized the model structure, thus increasing the classification speed of three different seeding rates by 121.12 %, 96.83 % and 103.09 %, respectively. At last, field planting tests were conducted, and the performance of mechanized transplanting when using graded density of HRMTBS was effectively improved. In the mechanized transplanting test using the graded density of the HRMTBS of 50 g/tray, 70 g/tray, and 90 g/tray, the average miss planting rate was decreased by 8.28 %, 4.67 %, and 3.19 %, respectively. The average qualification rates reached 93.33 %, 91.06 %, and 82.22 %, with an increase of 10.00 %, 14.89 %, and 8.89 % compared with mechanized transplanting using the un-graded HRMTBS. The average uniformity increased by 14.09 %, 12.04 %, and 12.83 %, respectively. The research results can provide a reference for the precise and intelligent mechanized transplanting of hybrid rice. [ABSTRACT FROM AUTHOR]

• A method for estimating leaf density of fruit-tree canopies based on the wind-excited audio. • Developed a spectral centroid attention based on the spectral-centroid distribution probability of wind-excited audio. • DCNN-based accurate estimation of leaf-densities of fruit-tree canopies. It is essential for precision air-assisted ground sprayer to accurately detect the leaf density of fruit trees in agriculture. However, it is difficult to achieve real-time measurement of canopy leaf densities by using conventional sensors (such as LIDAR and camera). In fact, the interaction between wind and canopies of different leaf densities may generate different excited-audios during air-assisted spraying. Therefore, this paper proposes a method for leaf-density detection of fruit-tree canopies based on the variance of wind-excited audios. First, the fused spectrogram (FSP) was constructed by the short-time fast Fourier transform (STFT), consisting of spectral features and the spectrogram of audio signals. Then, the estimation model between the FSP and leaf densities was developed by using deep convolutional neural network (DCNN), which was enhanced by a spectral centroid attention (SCA) based on the distribution function of spectral centroids. Finally, the test results showed that: (1) the developed model achieved a 96.93% accuracy and a 97.96% precision in leaf-density recognition, and (2) compared with unimproved DCNN model, the accuracy and precision of the developed model were increased about 7.85% and 8.47%, respectively, which indicated that this method could achieve the prediction of leaf-density based on wind-excited audios. The study is expected to provide a reference for leaf-density detection of fruit tree canopies and real-time control of precision spray. [ABSTRACT FROM AUTHOR]

• A New Framework for Effective Prediction of Long- and Short-Term Reservoir Levels. • A new method for extracting internal and external factors affecting water level. • A new method for obtaining an optimal time window with interpretability. • A new model that captures spatial properties in data. • Interpretation of prediction results using SHAP. The fluctuation of water levels in agricultural irrigation reservoirs is extremely important for effective planning and allocation of irrigation water resources by agricultural producers, and can be predicted using time-series neural networks. However, existing time-series models fail to capture the spatial characteristics in the input data. Furthermore, there is a lack of comprehensive analysis and extraction of internal and external factors that effectively influence water level changes in the dataset prior to model training, and the methods for determining optimal window values lack interpretability and scientific validity. Therefore, we propose a deep time-series water level prediction framework based on internal and external influencing factors. The framework includes three key components: a Data Preprocessing method based on Internal and External influencing factors (DPIE), a Time-series Window Parameter Optimization method based on Fast Fourier Transform (FTOM), and a CNN-SSA-GRU deep time-series model (CSG), which improve predictive performance in terms of input data quality, optimal time-series window value, and model structure respectively. Comparative experiments and ablation studies conducted on the test set demonstrate that the collaborative combination of these three components significantly enhances the accuracy of water level prediction. Multi-step prediction comparison experiments prove the framework's significant advantage in predicting water levels for multiple days ahead. Additionally, this paper uses the SHAP tool to analyze the interpretability of the decision-making process in the model's predictive results. In summary, this framework can assist agricultural producers in more effectively planning and allocating irrigation water resources and also provides a thought exploration and reference for water level prediction problems in other aquatic environments. [ABSTRACT FROM AUTHOR]

• The kinematic model of the agricultural tracked robot considering the slips (slippages and slip-rotation) between the tracks and the soil is established. • An improved sliding mode observer is proposed to estimate the slip parameters of agricultural tracked robots during the driving process. • Based on the given control sequence and estimated slip parameters, the driving trajectory of the agricultural tracked robot is predicted for a period of time in the future. The trajectory prediction of tracked robots is the foundation and prerequisite for trajectory tracking and autonomous precise navigation. The kinematic model of the agricultural tracked robot considering the slips (slippages and slip-rotation) between the tracks and the soil is established by analyzing the slip and turning characteristics. The extended Kalman filter (EKF) method and the improved sliding mode observer (ISMO) method are respectively used to estimate the slip parameters of the agricultural tracked robot during the driving process. Subsequently, the driving trajectory of the agricultural tracked robot is predicted for a future time period, in combination with the provided control sequence. Finally, simulation and experimental results show that the proposed trajectory prediction method for agricultural tracked robots, which integrates slip parameter estimation, significantly reduces trajectory prediction errors. Moreover, the proposed ISMO method outperforms the traditional EKF method in terms of slip parameter estimation and driving trajectory prediction. The research in this paper provides theoretical guidance for trajectory planning and tracking control of agricultural tracked robots, and has broad application prospects. [ABSTRACT FROM AUTHOR]

Pesticide spraying is one of the important measures to protect crops from pests and diseases in agricultural environment. In this paper, a precision multivariable spraying robot with swing fan structure is designed. The robot has the characteristics of variable flow rate, variable air volume, adjustable droplet diameter and adjustable spray direction. The function of changing spray direction is realized by crank rocker mechanism, and its swing angle reaches up to 30°. Target plants are extracted from the orchard environment point cloud by an improved DBSCAN(Density Based Spatial Clustering of Applications with Noise) algorithm. A multivariable spray model was proposed for controlling each spray unit and air-assisted unit according to the extracted target plant information. Further more, the robot can change the droplet diameter by adjusting the rotating speed of the centrifugal nozzle. The experiment results show that the multivariable spraying robot reduces the spray dosage by 83% compared with conventional spray and the spray effect is improved. • A multivariable spraying robot is designed to enhance spray efficiency. • The multivariate spray model mainly contains variable flow rate, variable air volume, and adjustable droplet diameter model. • The spraying robot can reduce the pesticide by 83% compared with conventional spraying. [ABSTRACT FROM AUTHOR]

The worldwide citrus market has been impacted by various factors in recent years, including population growth, phytosanitary diseases, high costs of agricultural inputs, and diminishing planting areas. As a consequence, producers in this sector have attempted to find tools to support strategic planting decisions, and thus meet international contract demands. This paper proposes an optimization tool for supporting the strategic planning of planting decisions in citriculture, based on mathematical models and algorithms that address real-world requirements. The motivation for this study stems from our collaboration with one of the world's largest orange juice producers. We consider specific characteristics of the citrus business, estimates for productivity and eradication, and desired balance levels for orange varieties and plant age groups. To the best of our knowledge, there are no previous studies proposing optimization approaches that explore these unique characteristics of citrus strategic planting. We validate the effectiveness of the proposed approach through computational experiments using realistic instances based on the company's data. The results show that our approach provides effective support to decision making and can significantly increase fruit box production over a 30-year planning horizon while, most importantly, satisfying all the company's requirements on varietal and age balance as well as planting and eradication control. • Strategic planning tool for in-citriculture planting. • Mathematical model and algorithms to support decision-making. • Tested by one the world's biggest producers of orange juice. • Optimized plan considering a horizon of 30 years. • Maximizes production while ensuring the desired age and varietal balance. [ABSTRACT FROM AUTHOR]

• Novel U-LSTM and FADA-A modules are introduced into FieldSeg-DA framework. • U-LSTM merges spatiotemporal features in multi-resolution satellite imagery. • FADA-A aligns cross-domain features with guidance of land cover product. • U-LSTM and FADA-A enhance temporal and spatial transferability respectively. Deep learning has become the leading technique for precisely extracting individual arable fields (IAFs) from high-resolution remote sensing images. Maintaining transferability remains a major concern for deep learning methodologies due to the significant cost of acquiring labelled samples for network training. FieldSeg-DA, introduced by Liu et al. (2022) , enhanced network transferability by introducing a finely tuned adversarial domain adaptation module (FADA). An improved version of the FieldSeg-DA framework, FieldSeg-DA2.0, which further enhances transferability through incorporating multisource remote sensing and land cover data is introduced in this paper. First, we introduce a spatiotemporal fusion module, U-LSTM, to extract the IAF extent by merging textural information from the high-resolution image (Gaofen-2) and phenological information from the coarse time-series data (Sentinel-2). Incorporating phenological information mitigates the risk of overfitting associated with the high-resolution imagery acquired in specific seasons, thereby improving the temporal transferability of FieldSeg-DA2.0 compared to FieldSeg-DA. Second, we introduce a novel fine-grained adversarial domain adaptation module with ancillary data (FADA-A) to enhance spatial transferability. FADA-A incorporates prior knowledge from the Dynamic World (DW) land cover dataset to guide adversarial training in the standard FADA, thereby enhancing the robustness of domain adaptation across diverse geographic regions. We evaluate the performance of FieldSeg-DA2.0 through various spatial and temporal transfer experiments utilizing GaoFen-2 and Sentinel-2 data. The results illustrate that the cross-domain performance of FieldSeg-DA2.0 is significantly better than that of the original FieldSeg-DA, highlighting its robust spatiotemporal transferability. FieldSeg-DA2.0 can accurately delineate IAFs across varied regions and seasons without requiring additional training samples, illustrating its considerable potential for large-scale IAF extraction. [ABSTRACT FROM AUTHOR]

• A disturbance simulation model of tooth-flexible body residual film-soil based on DEM-MBD was established. • The flow characteristics of film-soil was studied using DEM-MBD simulation methods. • A film picking test bench was established to verify the reliability of the simulation model. • The cross-sectional shape of soil was used to verify the accuracy of DEM simulation models. • The mechanical properties and tensile performance of the residual film were analyzed. The residual film in the cotton cultivation layer will seriously affect cotton cultivation, and mechanized recycling of residual film is currently the best way to solve this problem. To improve the recovery effect of the chain tooth tillage layer residual film recovery machine, a bionic elastic tooth was designed in this paper. Furthermore, DEM-MBD coupling multibody dynamics was used as the research method to study the bionic elastic tooth-residual film-soil interactive disturbance. A DEM-MBD simulation model for predicting the film lifting effect of a film lifting device has been established. And the DEM-MBD simulation model was used to optimize the design parameters of the film lifting device by studying the film lifting effect of the elastic teeth. The absolute value of the maximum curvature at the tip of the elastic tooth (A), the angle between the midline of the power wheel and the ground plane (B), and the spacing of the elastic teeth (C) were used as the influencing factors. The optimal solution parameters (A = 0.03, B = 54°, C = 83 mm) were obtained. The field test's film picking and disturbance rate was 91.27 %, the error with the simulation results (89.69 %) was 1.8 %, which verified the accuracy of the simulation results and the superiority of the optimized parameters. The recovery rates of residual film under soil compaction conditions of 600 kPa, 800 kPa, and 1000 kPa were 91.25 %, 90.94 %, and 90.89 % (Verified the strong adaptability of the device). This study can provide a theoretical research basis and test methods for the design optimization of touching parts. [ABSTRACT FROM AUTHOR]

• Eight new end-effectors for harvesting based on scissor mechanisms are proposed. • All the end-effectors can be driven by only one motor. • The end-effectors own good versatility due to their scalable structural features. • The end-effectors enable either direct separation or stem-holding separation. • Two approach orientations are designed to increase accessibility of fruits. Due to the escalating labour cost, manually harvesting of crops like apples, tomatoes and peppers, has become increasingly arduous and unsustainable. Automated harvesting employing contemporary robotic systems has emerged as a promising solution. However, a significant challenge in commercializing robot harvesting systems lies in developing effective robot end-effectors. This challenge is primarily attributed to factors such as high manufacturing costs, low picking efficiency, limited versatility, risks of fruit damage due to excessive gripping force, etc. To address these challenges, this paper proposes innovative end-effectors based on deployable mechanisms utilizing scissor mechanisms. A deployable scissor mechanism, distinguished by its expandability and foldability, can be compactly opened, and closed as an enclosable container to harvest one or multiple fruits simultaneously. Eight novel types of robot end-effector designs, each with one degree of freedom, are introduced for the first time for fruit harvesting based on deployable scissor mechanisms. These proposed end-effectors enable both direct and stem-holding separations for fruit prehension, offering two approaching orientations. Meanwhile, the proposed end-effectors can harvest different kinds of fruits of different sizes. To validate the concept, a prototype based on one of the proposed end-effector designs is manufactured by 3D printing to demonstrate its manufacturability and mobility. Simulations for automated harvesting are conducted, and harvesting efficiency analyses reveal that the robot harvesting system employing the proposed end-effector can enhance efficiency by 40% compared to the traditional approach. These newly designed end-effectors can also be applied in extended tasks such as picking sea cucumbers. [ABSTRACT FROM AUTHOR]

The past decade has witnessed the rapid development and adoption of machine and deep learning (ML & DL) methodologies in agricultural systems, showcased by great successes in applications such as smart crop management, smart plant breeding, smart livestock farming, precision aquaculture farming, and agricultural robotics. However, these conventional ML/DL models have certain limitations: they heavily rely on large, costly-to-acquire labeled datasets for training, require specialized expertise for development and maintenance, and are mostly tailored for specific tasks, thus lacking generalizability. Recently, large pre-trained models, also known as foundation models (FMs), have demonstrated remarkable successes in language, vision, and decision-making tasks across various domains. These models are trained on a vast amount of data from multiple domains and modalities. Once trained, they can accomplish versatile tasks with just minor fine-tuning and minimal task-specific labeled data. Despite their proven effectiveness and huge potential, there has been little exploration of applying FMs to agriculture artificial intelligence (AI). Therefore, this study aims to explore the potential of FMs in the field of smart agriculture. In particular, conceptual tools and technical background are presented to facilitate the understanding of the problem space and uncover new research directions in this field. To this end, recent FMs in the general computer science (CS) domain are reviewed, and the models are categorized into four categories: language FMs, vision FMs, multimodal FMs, and reinforcement learning FMs. Subsequently, the process of developing agriculture FMs (AFMs) is outlined and their potential applications in smart agriculture are discussed. In addition, the unique challenges and risks associated with developing AFMs are discussed, including model training, validation, and deployment. Through this study, the advancement of AI in agriculture is explored by introducing AFMs as a promising paradigm that can significantly mitigate the reliance on extensive labeled datasets and enhance the efficiency, effectiveness, and generalization of agricultural AI systems. To facilitate further research, a well-classified and actively updated list of papers on AFMs is organized and accessible at https://github.com/JiajiaLi04/Agriculture-Foundation-Models. • Basics of large language and foundation models. • Review of potential applications of large language and foundation models in agriculture. • Outline challenges and opportunities. [ABSTRACT FROM AUTHOR]

• The Multi-FVC model was developed based on PROSAIL model and Beer-Lambert law to retrieve FVC. • Multi-FVC model was more potential in estimating FVC and suppressing the influence of soil. • Soil effect caused by the second axis in the soil spectral line could be solved by NDVI and EVI. • Multi-FVC model could realize the conversion of FVC retrieval results at any two view zenith angles. Fractional vegetation cover (FVC) plays an important role in spectral unmixing, crop growth monitoring, crop light interception calculation, and yield estimation. However, the spectral reflectance would change with view zenith angles (VZAs), and retrieved FVC is also affected by VZAs. Therefore, in this paper, the observed multi-angular (±45°, ±30°, 0°) spectral datasets with different crops (wheat and rice), along with simulated spectral dataset, were used to explore method of correcting the influence of angle effect in FVC retrieval. Firstly, a simulated dataset of multi-angular hyperspectral data and directional FVC were constructed using the PROSAIL model, and a single-angular FVC retrieval model (Sin-FVC) was established based on the Gaussian process regression (GPR) algorithm. Then, the single-angular FVC (i.e., FVC θ) was converted into vertical FVC (FVC θv) based on the Beer-Lambert law. Finally, a Multi-FVC model was developed, which summed outputs of the Sin-FVC model (FVC 0° , FVC θv) according to their respective weights (i.e., FVC correct), to correct the angle effect. Using the pixel bisection model (FVC VI) as a comparison, different FVC retrieval models (Sin-FVC, Multi-FVC, and FVC VI) were compared and evaluated in terms of FVC retrieval accuracy, ability in weakening soil background, and LNC retrieval accuracy after spectral decomposition using different retrieved FVC (FVC 0° , FVC -30°v , FVC -45°v , FVC correct). The results showed that FVC correct had the highest FVC retrieval accuracy (RRMSE = 8.5 %) compared with FVC retrieved from other models (FVC VI , FVC 0° , FVC -30°v and FVC -45°v). Meanwhile, when using the reflectance at NIR band in the black soil background as the baseline, after decoupling mixing spectra based on FVC correct , the relative offset (i.e., RO correct) of each vegetation index (NDVI, EVI, SAVI, OSAVI) was minimum (least-RO correct = 0.42 %). That is, the soil background was effectively suppressed by FVC correct spectral decomposition, along with the highest LNC retrieval accuracy, with an RRMSE of 14.2 %. [ABSTRACT FROM AUTHOR]

• A convolutional neural network-based method is proposed for detecting wheat diseases characterized by arbitrary orientation and significant aspect ratio variations. • An auto-adjustment label assignment scheme based on the similarity of aspect ratio between the sample and object is proposed to assign high-potential positive samples. • A localization potential assessment scheme based on location distance is proposed to evaluate each positive sample. • A wheat disease dataset WDF2023 based on oriented annotation is established. The frequent occurrence of wheat diseases seriously affects the quality and yield of wheat. Thus, the accurate detection of wheat diseases is highly desired in the field of agricultural information. However, for wheat disease with arbitrary-oriented and aspect ratio varies greatly, existing deep learning-based methods adopt defined IoU threshold to assign label and do not consider the differences in localization potential of selected positive samples, resulting in some objects with large aspect ratios could not match enough high potential positive samples. In this paper, we put forward a convolutional neural network-based method for detecting wheat diseases characterized by arbitrary orientation and significant aspect ratio variations. First, we design an auto-adjustment label assignment scheme based on the similarity of aspect ratio between the sample and object to assign high-potential positive samples. Then, a localization potential assessment scheme is proposed to evaluate each positive sample. Finally, we construct a dataset of wheat disease in field (WDF2023) based on oriented annotation. We evaluate the effectiveness of our proposed method and eight oriented object detection detectors. The experimental results showcase that the proposed method attains an mAP of 60.8% and an mRecall of 73.8% on the WDF2023 dataset, surpassing existing advanced oriented object detection detectors. Notably, when contrasted with conventional horizontal object detection detectors, our method demonstrates superior performance in precisely localizing disease regions. [ABSTRACT FROM AUTHOR]

Apple plays a vitally important role in human life and is considered one of the most nutritious fruits. However, the quality and production of the apple industry are seriously restricted by apple leaf diseases and the disease lesions are hard to detect because they often have various scales and deformable geometry. To solve the above problem, this paper proposed a novel Multi-scale Constrained Deformable Convolution Network(MCDCNet), which takes advantage of multi-branch convolution and deformable convolution. Firstly, the novel two-branch convolution network is presented to enhance the discriminatory ability of models for extracting different scales of apple leaf disease. Secondly, different offset intervals are applied to the two kernels of the dual convolution channel separately, which makes the proposed model pay more attention to the deformable geometry features of the lesions and avoid extra weight parameters. Finally, a feature fusion module is constructed to achieve automatic detection of multi-scale apple leaf disease, which combines the output features from the dual convolution channels and performs dimensional operations on the channel dimensions of the feature map. Under the complex natural environment, the accuracy value of the proposed model can reach 66.8%, which is an improvement of 3.85% compared to the existing SOTA models. The experiment results established that MCDCNet has a better feature extraction capability and can efficiently and accurately detect 5 common apple leaf diseases in the natural environment. • The proposed MCDCNet can extract more reliable features of apple leaf diseases with various scales and geometry which effectively improve the discriminative ability of the network. • A novel Dual-constrained deformable convolution module is proposed to help the network get flexible receptive fields and help network dealing with apple leaf disease with various geometry and size. • A novel Feature fusion module is proposed to fuse outputs from dual branches which helps the MCDCNet automatically selecting appreciate geometry and scales of apple leaf disease. [ABSTRACT FROM AUTHOR]

The development of practical and robust automated diagnostic systems for identifying plant pests is crucial for efficient agricultural production. In this paper, we first investigate three key research questions (RQs) that have not been addressed thus far in the field of image-based plant pest identification. Based on the knowledge gained, we then develop an accurate, robust, and fast plant pest identification framework using 334K images comprising 78 combinations of four plant portions (the leaf front, leaf back, fruit, and flower of cucumber, tomato, strawberry, and eggplant) and 20 pest species captured at 27 farms. The results reveal the following. (1) For an appropriate evaluation of the model, the test data should not include images of the field from which the training images were collected, or other considerations to increase the diversity of the test set should be taken into account. (2) Pre-extraction of ROIs, such as leaves and fruits, helps to improve identification accuracy. (3) Integration of closely related species using the same control methods and cross-crop training methods for the same pests, are effective. Our two-stage plant pest identification framework, enabling ROI detection and convolutional neural network (CNN)-based identification, achieved a highly practical performance of 91.0% and 88.5% in mean accuracy and macro F1 score, respectively, for 12,223 instances of test data of 21 classes collected from unseen fields, where 25 classes of images from 318,971 samples were used for training; the average identification time was 476 ms/image. • Development of an accurate and practical framework for plant pest identification. • Evaluation of the model requires images, such as those collected elsewhere. • Foreground extraction significantly improves pest discrimination performance. • Class integration and cross-crop training boosts model performance. [ABSTRACT FROM AUTHOR]

• A sprinkler with an intermittent water-dispersing needle device was designed. • A precise optimization method of irrigation sprinkler was obtained. • Key reasons for successful sprinkler design were explored. Sprinklers with fixed water-dispersing needles (FWDN) are widely used in irrigation systems due to their high spray uniformity. However, they have a limited throw radius which increases investment costs. Hence, in this paper, we proposed an innovative sprinkler with Intermittent Water-Dispersing Needle (IWDN) to overcome this limitation. Key parameters of new sprinklers with IWDNs, such as sprinkler rotation speed n , measured in revolutions per minute (rpm); needle insertion frequency f , expressed in insertions per minute (ipm); needle insertion depth h , specified in millimeters (mm); and the distance between the needle and nozzle outlet l , also in millimeters (mm), were investigated. This was done by combining experiments, Random Forest and Non-Dominated Sorting Genetic Algorithm II (RF-NSGA II), and Computational Fluid Dynamics (CFD) simulations. The optimal configuration of the parameters was obtained through analysis, the hydraulic performance was optimized, and finally, the improvement mechanism was analyzed. The results indicated that the RF-NSGA II values matched experimental values, with throw radius errors below 2.82 % and combination uniformity errors under 4.97 %. Optimal matching ranges were n ∈ [0.54, 0.7] rpm, f ∈ [20, 24] ipm, h ∈ [0.43 d , 0.48 d ], and l ∈ [7 d , 9.5 d ], with d = 4.2 mm representing the nozzle outlet diameter. Within the optimal matching range, a set of parameters was selected to configure the IWDN sprinkler, yielding a maximum reduction in combination uniformity of 5.53 % compared to the sprinkler with FWDN. On the other hand, the throw radius increased by at least 29 %. In addition, a jet dispersion rate of approximately 6 % results in a satisfactory combination uniformity. The lower rate of jet velocity decreased and total entropy production in the IWDN accounted for the increased throw radius when the h ≥ 0.3 d. Additionally, the IWDN had lower air entrainment rates, fewer jet structures, and reduced jet dispersion rate, all of which contributed to a lower combination uniformity compared to FWDN. [ABSTRACT FROM AUTHOR]

AgroCounters is an open-source repository for counting objects in images in the agricultural domain by utilizing deep-learning algorithms. In this paper, we present the framework of AgroCounters, which integrates state-of-the-art deep learning models, including regression-based counting, detection-based counting, and density-estimation-based counting, to accurately count various agricultural objects, such as fruits, vegetables, and livestock, in single images. The framework utilizes transfer learning techniques to optimize model performance on the limited labeled data available in the agricultural domain. We provide an open-source implementation of AgroCounters, which includes a multitude of algorithms for counting applications and a toolbox that includes metrics, training data tools, visualizations, and a simple installation guide for several open-source implementations of counting methods. We evaluated the performance of AgroCounters on multiple agricultural datasets acquired from RGB sensors, including plant leaves, melons, wheat grains, cherry tomatoes, grapes, apple flowers, bananas (fruit and leaves), pears, and chickens. We compared the results of the various implemented methods over these datasets and showcased the most suitable solution for each. YOLOv5, the most recent of the compared object detectors, provided the best results on all the examined datasets, and there was no clear 'winner' between Faster-RCNN and RetinaNet. Based on the analyzed datasets, when higher accuracy is required, the direct regression network (DRN) should be used; for small datasets, multiple scale regression (MSR) gives superior results. Based on the developments, we proposed guidelines for developing deep-learning-based counting solutions for agricultural applications, focusing on solutions and best practices for the agricultural domain. Overall, AgroCounters presents a promising solution for automated counting in the agricultural domain, offering significant potential for reducing manual labor, improving crop management, and increasing productivity. • AgroCounters is a repository for counting agriculture objects using deep learning. • AgroCounters includes regression, detection, and density estimation algorithms. • AgroCounters provides toolboxes for metric calculation, training, and visualization. • AgroCounters publicizes 10 different datasets with annotations and trained models. • Present guidelines for developing deep learning-based counters in agriculture. [ABSTRACT FROM AUTHOR]

[Display omitted] • An apple dual-layer model based on velocity and bruise volume is verified. • The bruise volume of apple is visually extracted at different heights. • The velocity, force, stress, and energy of formation process of bruise is analysed. • The bruise susceptibility is compared based on FEM. To analyse the characteristics and internal mechanical changes of bruising formed on apples under different working conditions, this paper explored the formation mechanism of bruising using finite element method. A composite model of peel and flesh was established by reverse engineering, and the accuracy of the apple model was validated two aspects: drop velocity and bruise volume, with a maximum error of 2.51 %. The collision velocity, contact force, stress, and energy of apple bruising were analysed using commonly used contact materials (wood and rubber) at falling heights of 600, 900, 1200, 1500, 1800, and 2100 mm. The characteristics of apple bruising after formation were analysed and the susceptibility of apple bruising were explored. The findings suggest that as the height increased, so does the maximum stress and bruise susceptibility. At a height of 2100 mm, the apple's bruise susceptibility was measured at 24955.87 mm3 J−1 (wood) and 24710.18 mm3 J−1 (rubber), respectively. This study offers valuable insights into understanding the mechanism behind bruising in fruits, and has a guiding role in the control of bruising and the optimal design of related machinery. [ABSTRACT FROM AUTHOR]

• The performance of 2D LiDAR SLAM in simulated orchards is evaluated. • Three typical algorithms Hector, GMapping, and Cartographer are investigated. • A scalable terrain simulation model is built with adjustable roughness. • Evaluation metrics are localization/mapping errors and CPU/memory usage. • Adaptability to terrain roughness, LiDAR model, and orchard size is evaluated. Accurate localization is a prerequisite for developing autonomous mobile orchard robots. Simultaneous localization and mapping (SLAM) is an efficient technique for localizing robots in global navigation satellite system (GNSS)-denied scenarios. Light detection and ranging (LiDAR) is one of the most crucial sensors used for environment perception. Two-dimensional (2D) LiDAR SLAM has been successfully applied in various indoor scenes to construct flat maps and estimate robot trajectories. In this paper, the performances of three representative 2D LiDAR SLAM algorithms, namely, Hector, GMapping, and Cartographer, in semi-structured orchard environments simulated in Gazebo are analysed. A hierarchical terrain modelling method is proposed to generate scalable orchard terrain with adjustable roughness. The adaptability of the three algorithms to terrain roughness, LiDAR, and orchard size is evaluated in terms of the localization error, mapping error, CPU usage, and memory usage. The experimental results show that Cartographer has the highest location and mapping accuracy, followed by GMapping and Hector. However, Hector requires the least computational resources, followed by Cartographer and GMapping. In a 50 m × 50 m orchard with an elevation difference of 15 cm, Cartographer achieved a localization error of 8.14 cm and a mapping error of 8.43 cm at a 4 cm map resolution. In addition, Hector has the highest requirements for the maximum range and field-of-view (FOV) of LiDAR, and GMapping is most susceptible to severe uneven terrain conditions and has the worst scalability for large-scale orchards. [ABSTRACT FROM AUTHOR]

• YOLOv5-T can accurately identify maize tassels at different growth stages. • The average time for YOLOv5-T to detect maize tassels is 0.0237 s. • The mean Average Precision of maize tassels detection can reach 98.70 % Maize tassels is an important organ of maize plant, and has a very important impact on yield prediction and variety breeding. Therefore, realizing the efficient and accurate detection of maize tassels in the natural environment is the key to obtain maize phenotypes in large quantities and accurately predict maize yield. This study constructed a dataset of tassels from different growth stages, proposed a YOLOv5-Tassel (YOLOv5-T) maize tassels detection model based on UAV remote sensing platform, combined with attention mechanism, YOLOv5_l network, spatial pyramid pooling structure and multi-scale extraction advantages of Atrous convolution. The experimental results showed that the Average Precision (AP) of the model for maize tassels detection could reach 98.70 %, and the detection speed was 42.2f/s. And the method in this paper had strong robustness to changes in light intensity and maize tassels at different growth stages. It was feasible to use YOLOv5-T to detect large areas of maize tassels in real time, which provided a useful reference for the estimation of maize yield and the selection of maize varieties. [ABSTRACT FROM AUTHOR]