Early pest infestations can significantly reduce chili pepper yield and quality. Therefore, timely and accurate pest detection is essential for precision pest management. However, pest detection under practical cultivation conditions remains challenging because of small target sizes, dense distributions, partial occlusion, and complex background interference. To address these challenges, this study proposes HCFD-YOLOv8, a task-oriented lightweight visual detection framework based on YOLOv8n for pepper pest monitoring in complex agricultural environments. A novel Cross-Stage Partial Hybrid Spatial Attention (CSP-HSA) module is developed as the core feature enhancement component. It is designed to improve fine-grained pest feature extraction while reducing computational redundancy. CSP-HSA combines cross-stage partial feature reuse, heterogeneous convolution, lightweight spatial-channel attention, and channel shuffle to enhance discriminative representations of small and densely distributed pests. In addition, a Cross-Scale Context Fusion Module (CCFM) is introduced to improve information interaction between high-resolution spatial details and high-level semantic features across different scales. DyHead and Inner-MPDIoU are further incorporated as complementary components to enhance adaptive feature perception and bounding-box localization, especially for partially occluded and overlapping targets. Experimental results show that YOLOv8n achieves a Precision of 84.0% and an mAP50 of 87.1%, whereas HCFD-YOLOv8 improves these values to 90.9% and 92.0%, respectively, corresponding to increases of 6.9 and 4.9 percentage points. Meanwhile, the proposed framework reduces the number of parameters, model weight-file size, and FLOPs from 3.00 M to 2.43 M, from 5.97 MB to 4.96 MB, and from 7.5 G to 6.5 G, respectively. The model achieves an inference throughput of 39.40 FPS on the evaluated cloud-server platform. Comprehensive ablation experiments demonstrate that DyHead delivers the largest standalone improvement in detection accuracy, whereas CSP-HSA provides a more favorable trade-off between feature enhancement and computational efficiency. CCFM and Inner-MPDIoU further provide complementary improvements in cross-scale feature representation and bounding-box localization. These results indicate that HCFD-YOLOv8 achieves a favorable balance between detection accuracy and computational efficiency, demonstrating its potential for lightweight and accurate pepper pest monitoring under complex agricultural conditions.

Accurately detecting pepper pests in complex greenhouse environments is a crucial requirement for achieving a sequence of automated crop protection technologies, including precision spraying, real-time yield estimation, and intelligent pest management. However, the detection of small, fast-moving pests such as aphids, whiteflies, and Helicoverpa assulta remains a considerable challenge owing to their diminutive dimensions, comparable hue to the foliage, and frequent occlusion. Hence, we introduce a PDES-YOLO model, utilizing YOLOv8s as the foundation, for precise detection of pepper pests in greenhouse scenarios. First and foremost, a feature-focused diffusion pyramid network (FDPN) is introduced and integrated with a dimension-aware selective integration (DASI) to construct a novel dimension-selective fusion module (DSFM) structure, which asymptotically combines low- and high-level features via adaptive weighting, thus improving the model's capability to localize and distinguish small targets. Secondly, the initial C2f module in the backbone is substituted with our self-developed efficient multi-scale convolution extended module, C2f-EffiScale-X, to improve multi-scale feature representation and strengthen the sensitivity of network to varying pest sizes. Thirdly, the SPPF module is optimized by embedding a large separable kernel attention (LSKA) mechanism, forming the SPPF-LSKA module, which further refines spatial feature extraction and bolsters robustness against background clutter. Finally, to mitigate computational load while sustaining high-precision detection, the detection head is lightened by embedding the separated and enhancement attention module (SEAM) attention mechanism, which reduces model parameters and enhances resilience to overlapping occlusion. In the experimental section, a custom-built pepper pest dataset is implemented to quantify the model's accuracy. The improved PDES-YOLO model achieved precision, recall, and mean average precision (mAP@0.5) of 87.8%, 84.5%, and 87.8%, which are 7.3%, 6.9%, and 6.7% higher than YOLOv8s baseline, respectively. Moreover, through lightweight optimization, model parameters decreased by 21.6%, making the approach highly compatible for implementation on mobile and built-in devices. These experimental findings indicate that PDES-YOLO can efficiently detect pepper pests in intricate controlled environment, offering strong technical assistance for the development of smart pest surveillance and control systems.

To address the challenges of unmanned agricultural robots operating in unstructured environments, this study proposes an integrated control framework that synergizes a modified super twisting sliding mode controller with an adaptive data-driven super twisting extended state observer. First, the path tracking problem is decoupled into lateral deviation and heading angle tracking, resulting in a preview error mixed dynamic model. Second, a data-driven adaptive extended state observer is developed that simultaneously estimates the unknown control input gain and the lumped disturbance in real time. By exploiting both historical and current data through an integrated parameter learning law, the observer significantly accelerates convergence. Using these online estimates, a composite modified super twisting sliding mode controller is developed, and the stability of the closed-loop system has been rigorously verified. The key innovation of this approach lies in its model-free capability to concurrently estimate system lumped disturbance and unknown system parameters, thereby eliminating any requirement for prior knowledge of the model parameters. Finally, comparative high-fidelity simulations and field experiments are carried out to demonstrate the superiority of the proposed method.

To address the problems of ginger stalks being easily broken and incompletely cut during the clamp-pull ginger harvesting process, two ginger varieties, small yellow ginger and red ginger, were selected as experimental subjects. The basic biophysical parameters of the two varieties were measured, and the mechanical properties of their stalks were tested through shear, compression, tensile, and pull-out tests. The results showed that the mechanical properties of red ginger stalks were superior to those of small yellow ginger, and the stalks were less prone to breakage during harvesting. Moreover, the mechanical properties of the middle part of red ginger stalks were better than those of the upper and lower parts, making it more suitable as the clamping position in clamp-pull harvesting. The maximum shear strength of red ginger stalks was 1.42 MPa, the maximum compressive strength was 0.21 MPa, the maximum tensile strength was 4.31 MPa, and the average pull-out force was 68.45 N. This study provides a basis for low-damage ginger harvesting and for the selection of varieties suitable for mechanized harvesting.

Accurate monitoring of crop pests and diseases is essential due to their diversity, widespread impact, and increasing severity. However, RGB images lack spectral richness, and hyperspectral images suffer from resolution loss caused by long exposure times. Existing fusion methods often lose important features, reducing the accuracy and practicality of pest and disease monitoring. To address these issues, this paper proposes a SADAE-GAN model, which combines Deep Convolutional Generative Adversarial Networks (DCGAN), Self-Attention (SA), and Denoising Auto-Encoders (DAE) techniques. The model enhances both local and global feature extraction through a primary processing module and DAE in the generator, while the discriminator uses an SA attention mechanism to improve image reconstruction quality. Additionally, a novel perceptual gradient loss function and an Autoencoder and Continuous Wavelet Transform (Auto-CWT) image transformation method are introduced to optimize image generation and monitoring performance. Experimental results across multiple datasets demonstrate that the proposed SADAE-GAN outperforms existing popular fusion methods like CycleGAN, ResNet, SSRNET, InfoGAN, and QIS-GAN. For evaluation metrics SSIM, PSNR, MSE, and IS, the maximum performance improvement reached 20.7 %, 68.2 %, 39.7 %, and 33.0 %, respectively. In addition, ablation studies confirm the effectiveness of the proposed model, and robustness tests further reveal the model’s strong practical applicability. These results demonstrate that SADAE-GAN excels in fusion quality, robust feature preservation, and strong generalization capability, providing a promising and efficient solution for precise agricultural pest and disease monitoring.

Aphids are the main agricultural pests that affect the quality and yield of peppers in the greenhouse. Efficient early prediction of aphid occurrence is of great significance for the development of digitization and information technology in intelligent agriculture. Forecasting accuracy could be improved by the incorporation of feature interactions into pest forecasting. This study integrates multiple environmental factors to efficiently predict the number of aphids and the aphid strain rate in the greenhouse. We propose a two-level distributed multi-source information fusion approach, which integrates a one-dimensional convolutional neural network (1D CNN) and Long Short-Term Memory (LSTM). To enhance the accuracy of regional environmental parameters, a weighted average algorithm employs environmental sensor data in the first level of fusion. In the second fusion level, a heterogeneous sensor fusion algorithm allows for the integration of multi-source data to model the connection between environmental factors and aphid dynamics. Finally, the improved 1D CNN-LSTM fusion model and other models were tested to verify the effectiveness and robustness of the proposed model. The experimental results show that the total root mean square error of the proposed model is 1.503, which is obviously better than the other networks. In the test set, the total root mean square error of the model for predicting the aphid number and strain rate is 1.378 and 0.337, respectively, compared with existing network models such as 1D CNN, LSTM, and back propagation (BP). The experimental results show that the proposed model has obvious advantages for predicting the aphid number and strain rate. It provides a promising step forward in pest management, offering precise, environmentally friendly solutions that enhance crop yield and quality.

The operation of roller-type pepper harvesters involves striking and pulling the peppers, which may result in incomplete detachment from their stems or cause surface damage or breakage of the peppers. The quality of the harvested peppers is directly influenced by the forces applied during striking and pulling. Therefore, the physical and mechanical properties of peppers are crucial for determining the structural and dynamic parameters of the screw rollers. This study selected Round peppers and the 'Bo 15' line pepper as experimental subjects. Growth parameters such as main stem diameter and fruit diameter were measured. And mechanical properties including tensile, bending, compressive, and shear strength were tested. Results showed that: two pepper varieties had a moisture content of 90% ± 1%. The main stem diameters of Round pepper and 'Bo 15' line pepper were 10.196 ± 1.508 mm and 13.44 ± 0.769 mm. The average diameter of round pepper was greater than that of 'Bo 15' line pepper. For mechanical stress, the 'Bo 15' line pepper exhibited stronger resistance, and the mechanical properties were as follows: tensile strength was 0.83 MPa, bending strength was 0.58 MPa, radial compressive strength was 0.25 MPa, and shear strength was 0.28 MPa. This study provides a basis for the design of low damage harvesting device for fresh peppers and the selection of varieties suitable for mechanical harvesting.

Compared to traditional agricultural environments, facility-based agricultural environments are spatially compact and poorly signalized, making it difficult to realize accurate navigation operations. At the same time, the complexity of interfacing the different operational areas within the facility has resulted in unnecessary paths for the platform during regional transitions. To tackle these challenges, this study designs an autonomous navigation mobile platform suitable for working in a facility environment. Then, we implement a dynamic weighted heuristic function and an optimized search neighborhood direction for the global path planning in A* algorithm to improve the computational efficiency. The simulation results using MATLAB R2021a software show that the improved algorithm efficiency has increased by 30%. Furthermore, we use the Depth First Search (DFS) algorithm to reduce the total distance of articulation, smooth the path by optimizing the number of nodes, and verify the integration algorithm through simulation. Finally, field experiment is conducted and the results show that the computation time is reduced by 30%, the number of critical nodes is reduced by 50%, the operations coverage ratio and operations distance ratio are also increased by 7.9% and 8.6%, respectively, when applying the improved A* and DFS algorithm, which meets the requirements of the facility environment operations.

BACKGROUNDG protein-coupled receptors (GPCRs) are very promising as the targets of endogenous neuropeptides/neuromodulators that, upon binding to receptors, induce profound changes in insect physiology. The Methuselan/Methuselan-like subfamily of GPCRs is reported to be associated with longevity and stress resistance. A previous study showed the fungicide jingangmycin-induced expression of Mthl2 and enhanced stress resistance in Nilaparvata lugens. However, the other physiological functions of Mthl2 remain unelucidated.RESULTSThe Mthl2 was highly expressed before molting and decreased after that until the next ecdysis, showing a cyclical pattern related to molting behavior and predominantly distributed in cuticle-producing and reproductive tissues in N. lugens. Silencing Mthl2 by RNAi in nymphs disrupted the synthesis of 20E, caused downregulation of the 20E signaling-related genes, and further affected the transcription of cuticular proteins. Moreover, it led to the malformation of the integument structure and a declined emergence rate, whereas exogenous 20E could rescue the declined emergence rate caused by knockdown of Mthl2. Furthermore, depletion of Mthl2 through RNAi in the N. lugens nymphal stage influenced the development of the ovaries and fecundity in female adults. The soluble protein content in reproductive tissues, the protein and transcript levels of Vitellogenin (Vg) were significantly decreased after silencing of Mthl2, ultimately leading to a decline in the number of offspring with an obviously transgenerational consequence.CONCLUSIONThe current study revealed the physiological functions of Mthl2 in molting and fecundity of N. lugens, which can be used as an RNAi-based insecticide discovery to control this pest. (c) 2025 Society of Chemical Industry.

In this research, we present an adaptive fuzzy prescribed time second-order sliding mode (SOSM) control strategy designed to address a category of nonlinear system subject to asymmetric output restrictions. The proposed method employs fuzzy logic systems (FLS) to estimate the bounds of uncertainty, thereby relaxing the assumptions and alleviating chattering issues inherent in traditional SOSM methods. To effectively address asymmetric output restrictions, we utilize a robust approach known as the barrier Lyapunov function (BLF). Furthermore, the incorporation of a time-varying scaling function guarantees prescribed-time convergence of the closed-loop system while avoiding the computational singularity issue. By integrating these methodologies and introducing one power integrator technique, the prescribed-time adaptive fuzzy SOSM framework is established. Rigorous analysis based on Lyapunov stability theory substantiates the prescribed-time performance of the controlled system under the designed control scheme, satisfying asymmetric constraints. A notable innovation of this work is the assurance of stability within a prescribed time span, regardless of initial conditions. Finally, numerical simulation outcomes underscore the forcefulness of the designed method.

Highlights The spatial coordinate position of peppers is obtained by YOLOv5 algorithm fused with localization technique. The inverse kinematics for gathering arm is solved by disassembly and Euler-angle method. Pentadic polynomials are applied to realize the motion planning and control of robotic arm. The proposed method achieves an average gathering success rate about 60%. Abstract. To avoid mutation and interference with intelligent gathering, an autonomous gathering method is proposed in this paper. An improved YOLOv5 target detection algorithm is used to recognize peppers. The kinematics solution of the robotic arm is realized by using the disassembly method and the Euler-angle method. Simulating and modeling robotic arms is conducted by the Robotic Toolbox in Matlab. The robotic arm's spatial motion planning is controlled by the pentadic polynomial interpolation method. The feasibility of the inverse solution and planning is verified in Matlab. Different conditions’ gathering experiments are conducted on the gathering platform. Experiments show that this method is smooth and interference-free, with a high gathering success rate, realizing the accurate gathering of single, small fruits in complex environments. Keywords: Robot gathering, Robotic arm inverse kinematics, Target detection, Trajectory planning

Protecting crops from pests is a major issue in the current agricultural production system. The agricultural digital twin system, as an emerging product of modern agricultural development, can effectively achieve intelligent control of pest management systems. In response to the current problems of heavy use of pesticides in pest management and over-reliance on managers’ personal experience with pepper plants, this paper proposes a digital twin system that monitors changes in aphid populations, enabling timely and effective pest control interventions. The digital twin system is developed for pest management driven by data and model fusion. First, a digital twin framework is presented to manage insect pests in the whole process of crop growth. Then, a digital twin model is established to predict the number of pests based on the random forest algorithm optimized by the genetic algorithm; a pest control intervention based on a twin data search strategy is designed and the decision optimization of pest management is conducted. Finally, a case study is carried out to verify the feasibility of the system for the growth state of pepper and pepper pests. The experimental results show that the virtual and real interactive feedback of the pepper aphid management system is achieved. It can obtain prediction accuracy of 88.01% with the training set and prediction accuracy of 85.73% with the test set. The application of the prediction model to the decision-making objective function can improve economic efficiency by more than 20%. In addition, the proposed approach is superior to the manual regulatory method in pest management. This system prioritizes detecting population trends over precise species identification, providing a practical tool for integrated pest management (IPM).

To avoid mutation and interference with intelligent gathering, an autonomous gathering method is proposed in this paper. An improved YOLOv5 target detection algorithm is used to recognize peppers. The kinematics solution of the robotic arm is realized by using the disassembly method and the Euler-angle method. Simulating and modeling robotic arms is conducted by the Robotic Toolbox in Matlab. The robotic arm's spatial motion planning is controlled by the pentadic polynomial interpolation method. The feasibility of the inverse solution and planning is verified in Matlab. Different conditions' gathering experiments are conducted on the gathering platform. Experiments show that this method is smooth and interference-free, with a high gathering success rate, realizing the accurate gathering of single, small fruits in complex environments.

Hydrogen is an efficient and environmentally friendly energy source, and the research of hydrogen storage technology is of extreme importance. Solid-state hydrogen storage technology using metal hydrides as carriers has good application prospects. This paper aims to optimize the heat transfer resistance and hydrogen absorption kinetics encountered in practical applications of reactors with LaNi5 hydrogen storage materials. A novel fin structure is proposed to optimize the performance of a U-tube heat exchanger type reactor. The numerical simulation models of hydrogen absorption reaction in a single U-tube heat exchanger reactor and a reactor with fins installed in a straight section of the U-tube are established. The comparison results showed that the auxiliary heat dissipation of heat transfer fins accelerated the speed of hydrogen absorption reaction to a certain extent, but did not achieve the expected effect. Based on this situation, an additional curved fin was added to the U-tube bend section. The results showed that the addition of curved fins can overcome the inherent defects of the U-tube heat exchanger, significantly improve the heat dissipation capacity of the reactor, and the reaction rate approximately doubled. To further enhance the performance of the reactor, the operating conditions of the heat exchanger and reactor were discussed, including HTF flow rate, inlet temperature, as well as hydrogen absorption pressure. The results indicate that there is an optimal inlet flow rate of 2 m/s for HTF. The increase in hydrogen absorption pressure and the decrease in HTF's inlet temperature can effectively enhance the heat transfer and H2 absorption rate of the reactor.

The seedling retrieval mechanism is a crucial component of fully automatic transplanting machines, significantly influencing the quality, reliability, and efficiency of the transplanting process. Nonetheless, the existing seedling retrieval mechanisms in current transplanting machines exhibit several deficiencies, including substantial damage to seedlings and inadequate retrieval accuracy. To overcome these challenges, we propose an integrated approach combining pneumatic and mechanical techniques to further improve performance. By employing a lower thimble elevation and clamping mechanism, alongside a mathematical model based on the seedling removal process, this method ensures precise seedling extraction and minimizes damage to the root system and substrate. The novelty of this study lies in its ability to reduce the adhesion between seedlings and the holes of the plug plate, thereby minimizing non-destructive extraction of the seedlings and preserving the integrity of the matrix, which is essential for ensuring healthy seedling growth. Moreover, the optimization of the seedling retrieval trajectory enhances the accuracy of the seedling retrieval mechanism while also meeting the requisite speed requirements. Experimental results indicate that at a rate of 72 seedlings per minute, the extraction success rate reached 94.90%, and the casting success rate was 98.53%. The seedling injury rate was only 1.95%, resulting in an overall success rate of 91.69%. These findings confirm that the device meets operational efficiency requirements and delivers effective performance.

Mechanised watercress harvesting involves clamping and cutting its stalks, which can result in their incomplete breaking and crushing. The harvest quality is directly affected by the force used to clamp and cut the watercress stalks. Therefore, studying the physical and mechanical properties of the stalks is important to accurately calculate the force that needs to be applied. Herein, the microstructure of watercress stalk sections with and without nodes was observed using scanning electron microscopy. Moreover, the basic physical properties, such as the total length, internode outer diameter, internode inner diameter and water content, of watercress stalks were measured. Additionally, the mechanical properties of watercress stalks at different positions were measured using four modes: tension, compression, shear and bending. Results revealed that watercress stalks with nodes exhibited a more pronounced medullary cavity and a greater number of internal vascular bundles. The lower section of the watercress stalks was considerably more resistant to mechanical stress than the rest of the stalk. Further, in terms of resisting load, the stalks with nodes were stronger than those without nodes. This study provides useful information for efficient watercress harvesting.

Highlights An improved YOLOv5s deep learning model was used to identify peppers in complex background. The deep-level features on 3D (O-XYZ) coordinate of peppers were extracted using RealSense depth camera. An image database set of pepper in different scenes was established. A pepper recognition and location system were constructed based on improved YOLOv5s network. The proposed method achieved a mean average precision of 95.6% and minimum depth error of 0.001 m. Abstract. In order to investigate the impact of different scenes on the recognition performance and obtain the location information of picking targets, the recognition and location system based on improved YOLOv5s network and RealSense depth camera was constructed in this study. An image database in different scenes was established including light intensity, occlusion and overlap degree of pepper. An improved YOLOv5s deep learning model with bidirectional feature pyramid network (BiFPN) was used for the deep feature extraction and high-precision detection of pepper, and the effects of different scenes on recognition accuracy of the model were studied. The results showed that mean average precision (mAP) of YOLOv5s model reached 0.956, which was respectively 6.1%, 9.3%, 44.4%, and 8.2% higher than that of YOLOv4, YOLOv3, YOLOv2, and Faster R-CNN model. The model had good robustness under daytime and evening scenes with the mAP value higher than 0.9. The detection accuracy of the model in the leaf occlusion scenes was better than that of fruit overlap. The detection error was 0.001m which could not affect the picking positioning precision when the Z value of three-dimensional coordinates (O-XYZ) of pepper was 0.2 m. The improved algorithm can accurately recognize and extract three-dimensional coordinates of pepper, which reduces the calculations by eliminating lots of duplicate and redundant prediction boxes and provides a reference for trajectory planning of pepper picking operation. Keywords: Different scenes, Pepper recognition and location, Picking operation, YOLOv5s.

Due to changes in light intensity, varying degrees of aphid aggregation, and small scales in the climate chamber environment, accurately identifying and counting aphids remains a challenge. In this paper, an improved YOLOv5 aphid detection model based on CNN is proposed to address aphid recognition and counting. First, to reduce the overfitting problem of insufficient data, the proposed YOLOv5 model uses an image enhancement method combining Mosaic and GridMask to expand the aphid dataset. Second, a convolutional block attention mechanism (CBAM) is proposed in the backbone layer to improve the recognition accuracy of aphid small targets. Subsequently, the feature fusion method of bi-directional feature pyramid network (BiFPN) is employed to enhance the YOLOv5 neck, further improving the recognition accuracy and speed of aphids; in addition, a Transformer structure is introduced in front of the detection head to investigate the impact of aphid aggregation and light intensity on recognition accuracy. Experiments have shown that, through the fusion of the proposed methods, the model recognition accuracy and recall rate can reach 99.1%, the value mAP@0.5 can reach 99.3%, and the inference time can reach 9.4 ms, which is significantly better than other YOLO series networks. Moreover, it has strong robustness in actual recognition tasks and can provide a reference for pest prevention and control in climate chambers.

PurposeSimultaneous localization and map building (SLAM), as a state estimation problem, is a prerequisite for solving the problem of autonomous vehicle motion in unknown environments. Existing algorithms are based on laser or visual odometry; however, the lidar sensing range is small, the amount of data features is small, the camera is vulnerable to external conditions and the localization and map building cannot be performed stably and accurately using a single sensor. This paper aims to propose a laser three dimensions tightly coupled map building method that incorporates visual information, and uses laser point cloud information and image information to complement each other to improve the overall performance of the algorithm. Design/methodology/approachThe visual feature points are first matched at the front end of the method, and the mismatched point pairs are removed using the bidirectional random sample consensus (RANSAC) algorithm. The laser point cloud is then used to obtain its depth information, while the two types of feature points are fed into the pose estimation module for a tightly coupled local bundle adjustment solution using a heuristic simulated annealing algorithm. Finally, the visual bag-of-words model is fused in the laser point cloud information to establish a threshold to construct a loopback framework to further reduce the cumulative drift error of the system over time. FindingsExperiments on publicly available data sets show that the proposed method in this paper can match its real trajectory well. For various scenes, the map can be constructed by using the complementary laser and vision sensors, with high accuracy and robustness. At the same time, the method is verified in a real environment using an autonomous walking acquisition platform, and the system loaded with the method can run well for a long time and take into account the environmental adaptability of multiple scenes. Originality/valueA multi-sensor data tight coupling method is proposed to fuse laser and vision information for optimal solution of the positional attitude. A bidirectional RANSAC algorithm is used for the removal of visual mismatched point pairs. Further, oriented fast and rotated brief feature points are used to build a bag-of-words model and construct a real-time loopback framework to reduce error accumulation. According to the experimental validation results, the accuracy and robustness of the single-sensor SLAM algorithm can be improved.