Explore the vast range of titles available.

The identification of sweet corn seed vitality is an essential criterion for selecting high-quality varieties. In this research, a combination of hyperspectral imaging technique and diverse deep learning algorithms has been utilized to identify different vitality grades of sweet corn seeds. First, the hyperspectral data of 496 seeds, including four viability-grade seeds, are extracted and preprocessed. Then, support vector machine (SVM) and extreme learning machine (ELM) are used to construct the classification models. Finally, the one-dimensional convolutional neural networks (1DCNN), one-dimensional long short-term memory (1DLSTM), the CNN combined with the LSTM (CNN-LSTM), and the proposed firefly algorithm (FA) optimized CNN-LSTM (FA-CNN-LSTM) are utilized to distinguish spectral images of sweet corn seeds viability grade. The findings from the experimental analysis indicate that the deep learning models exhibit a significant advantage over traditional machine learning approaches in the discrimination of seed vitality levels, boasting a classification accuracy exceeding 94.26% in test datasets and achieving an accuracy improvement of at least 3% compared to the best-performing machine learning model. Moreover, the performance of the FA-CNN-LSTM model proposed in this study demonstrated a slight superiority over the other three models. Besides, the FA-CNN-LSTM achieved a classification accuracy of 97.23%, representing a significant improvement of 2.97% compared to the lowest-performing CNN and a 1.49% enhancement over the CNN-LSTM. In summary, this study reveals the potential of integrating deep learning with hyperspectral imaging as a promising alternative for discriminating sweet corn seed vitality grade, showcasing its value in agricultural research and cultivar breeding.

Multilevel thresholding image segmentation has attracted a lot of attention in the last several years since it has plenty of applications. The traditional exhaustive search methods are efficient for bi-level thresholding. However, they are time-consuming when extended to multilevel thresholding. To tackle this problem, a novel adaptive firefly algorithm (AFA) for multilevel thresholding using the minimum cross-entropy as its objective function has been proposed in this paper. The performance of the proposed algorithm has been examined on a set of benchmark images using various numbers of thresholds and has been compared with five different firefly variant algorithms. The experimental results indicated that the proposed algorithm outperformed the other five algorithms in terms of image segmentation quality, accuracy, and computation time.

This study investigates the impact of varying side wind velocities and nozzle inclination angles on droplet penetration during plant protection spraying operations, focusing on citrus trees. Experiments were conducted across four wind speed levels (0, 1, 2, 3 m/s) perpendicular to the nozzle direction and seven nozzle inclination levels (0°, 8°, 15°, 23°, 30°, 38°, 45°) to evaluate droplet distribution under different spraying parameters. A baseline condition with 0 m/s wind speed and a 0° nozzle angle served as the control. Utilizing Computational Fluid Dynamics (CFD) and regression analysis techniques in conjunction with field trials, the droplet penetration was analyzed. Results indicate that at constant wind speeds, adjusting the nozzle inclination angle against the direction of the side wind can significantly enhance droplet deposition in the canopy, with a 23° inclination providing the optimal increase in deposition volume, averaging a change of +16.705 μL/cm2. Multivariate nonlinear regression analysis revealed that both wind speed and nozzle inclination angle significantly affect the droplet penetration ratio, demonstrating a correlation between these factors, with wind speed exerting a greater impact than nozzle angle. Increasing the nozzle inclination angle at higher wind speeds improves the penetration ratio, with the optimal parameters being a 23° angle and 3 m/s wind speed, showing a 12.6% improvement over the control. The model fitted for the impact of nozzle angle and wind speed on droplet penetration was validated through field experiments, identifying optimal angles for enhancing penetration at wind speeds of 1, 2, and 3 m/s as 8°, 17°, and 25°, respectively. This research provides insights for improving droplet penetration techniques in plant protection operations.

UAVs for crop protection hold significant potential for application in mountainous orchard areas in China. However, certain issues pertaining to UAV spraying need to be addressed for further technological advancement, aimed at enhancing crop protection efficiency and reducing pesticide usage. These challenges include the potential for droplet drift, limited capacity for pesticide solution. Consequently, efforts are required to overcome these limitations and optimize UAV spraying technology. In order to balance high deposition and low drift in plant protection UAV spraying, this study proposes a plant protection UAV spraying method. In order to study the operational effects of this spraying method, this study conducted a UAV spray and grid impact test to investigate the effects of different operational parameters on droplet deposition and drift. Meanwhile, a spray model was constructed using machine learning techniques to predict the spraying effect of this method. This study investigated the droplet deposition rate and downwind drift rate on three types of citrus trees: traditional densely planted trees, dwarf trees, and hedged trees, considering different particle sizes and UAV flight altitudes. Analyzing the effect of increasing the grid on droplet coverage and deposition density for different tree forms. The findings demonstrated a significantly improved droplet deposition rate on dwarf and hedged citrus trees compared to traditional densely planted trees and adopting a fixed-height grid increased droplet coverage and deposition density for both the densely planted and trellised citrus trees, but had the opposite effect on dwarfed citrus trees. When using the grid system. Among the factors examined, the height of the sampling point exhibited the greatest influence on the droplet deposition rate, whereas UAV flight height and droplet particle size had no significant impact. The distance in relation to wind direction had the most substantial effect on droplet drift rate. In terms of predicting droplet drift rate, the BP neural network performed inadequately with a coefficient of determination of 0.88. Conversely, REGRESS, ELM, and RBFNN yielded similar and notably superior results with a coefficient of determination greater than 0.95. Notably, ELM demonstrated the smallest root mean square error.

Citrus pests and diseases severely impact fruit yield and quality. However, existing object detection models face limitations in complex backgrounds, target occlusion, and small target recognition, and they struggle to be efficiently deployed on resource-constrained devices. To address these issues, this study proposes a lightweight pest and disease detection model, YOLOv11-RDTNet, based on the improved YOLOv11n. This model integrates multi-scale features and attention mechanisms to enhance recognition performance in complex scenarios, while adopting a lightweight design to reduce computational costs and improve deployment adaptability. The model introduces three key enhancement features: First, shallow RFD (SRFD) and deep RFD (DRFD) downsampling modules replace traditional convolution modules, improving image feature extraction accuracy and robustness. Second, the Dynamic Group Shuffle Transformer (DGST) module replaces the original C3k2 module, reducing the model’s parameter count and computational demand, further enhancing efficiency and performance. Lastly, the lightweight Task Align Dynamic Detection Head (TADDH) replaces the original detection head, significantly reducing the parameter count and improving accuracy in small-object detection. After processing the collected images, we obtained 1382 images and constructed a dataset containing five types of citrus pests and diseases: anthracnose, canker, yellow vein disease, coal pollution disease, and leaf miner moth. We applied data augmentation on the dataset and conducted experimental validation. Experimental results showed that the YOLOv11-RDTNet model had a parameter count of 1.54 MB, an mAP50 of 87.0%, and a model size of 3.4 MB. Compared to the original YOLOv11 model, the YOLOv11-RDTNet model reduced the parameter count by 40.3%, improved mAP50 by 4.8%, and reduced the model size from 5.5 MB to 3.4 MB. This model not only improved detection accuracy and reduced computational load but also achieved a balance in performance, size, and speed, making it more suitable for deployment on mobile devices. Additionally, the research findings provided an effective tool for citrus pest and disease detection with small sample sizes, offering valuable insights for citrus pest and disease detection in agricultural practices.

This study examines the impact of canopy density, side wind speed, nozzle tilt angle, and droplet size on droplet penetration during plant protection spraying operations. Experiments conducted in citrus orchards evaluated how side wind speed and nozzle tilt angle influence droplet penetration across various canopy densities. A Phase Doppler Analyzer (PDA) was used to assess droplet size variations under different nozzle tilt angles and side wind speeds, yielding a multiple linear regression equation (R 2 = 0.866) that links nozzle tilt angle and side wind speed with droplet size. Results showed that droplet size decreases with increasing nozzle tilt angle at a constant crosswind speed. Further experiments investigated the effects of droplet size and canopy leaf area density on droplet penetration, involving three canopy leaf area densities, four wind speeds, and six nozzle tilt angles. Droplet deposition and canopy coverage were measured under various spraying parameters, with conventional operations (0° nozzle tilt and orthogonal wind speeds) serving as controls. The study found that adjusting nozzle tilt angle and wind speed enhances droplet penetration in different canopy structures. Optimal parameters varied with leaf area density (LAD): an 18° tilt angle and 3 m/s wind speed for a LAD of 5.94 m 3 /m 3 , a 45° tilt angle and 2 m/s wind speed for a LAD of 8.47 m 2 /m 3 , and a 36° tilt angle and 3 m/s wind speed for a LAD of 11.12 m 2 /m 3 . At 1 m/s, droplet deposition followed a downward parabolic trend with changes in nozzle tilt angle, whereas at 2 m/s, deposition followed an upward parabolic trend. At a side wind speed of 3 m/s, droplet deposition remained unchanged with nozzle tilt angle but decreased with increasing canopy density. Nonlinear regression analysis indicated that leaf area density had a greater impact on deposition differences than droplet size, with droplet penetration decreasing as leaf area density increased. This study provides a reference for enhancing fog droplet penetration techniques in plant protection operations, offering practical guidelines for optimizing spraying conditions and improving pesticide use efficiency in different canopy structures.

The accurate and rapid acquisition of fruit tree canopy parameters is fundamental for achieving precision operations in orchard robotics, including accurate spraying and precise fertilization. In response to the issue of inaccurate citrus tree canopy segmentation in complex orchard backgrounds, this paper proposes an improved DeepLabv3+ model for fruit tree canopy segmentation, facilitating canopy parameter calculation. The model takes the RGB-D (Red, Green, Blue, Depth) image segmented canopy foreground as input, introducing Dilated Spatial Convolution in Atrous Spatial Pyramid Pooling to reduce computational load and integrating Convolutional Block Attention Module and Coordinate Attention for enhanced edge feature extraction. MobileNetV3-Small is utilized as the backbone network, making the model suitable for embedded platforms. A citrus tree canopy image dataset was collected from two orchards in distinct regions. Data from Orchard A was divided into training, validation, and test set A, while data from Orchard B was designated as test set B, collectively employed for model training and testing. The model achieves a detection speed of 32.69 FPS on Jetson Xavier NX, which is six times faster than the traditional DeepLabv3+. On test set A, the mIoU is 95.62%, and on test set B, the mIoU is 92.29%, showing a 1.12% improvement over the traditional DeepLabv3+. These results demonstrate the outstanding performance of the improved DeepLabv3+ model in segmenting fruit tree canopies under different conditions, thus enabling precise spraying by orchard spraying robots.

Target detection technology and variable-rate spraying technology are key technologies for achieving precise and efficient pesticide application. To address the issues of low efficiency and high working environment requirements in detecting tree information during variable spraying in orchards, this study has designed a variable spraying control system. The system employed a Kinect sensor to real-time detect the canopy volume of citrus trees and adjusted the duty cycle of solenoid valves by pulse width modulation to control the pesticide application. A canopy volume calculation method was proposed, and precision tests for volume detection were conducted, with a maximum relative error of 10.54% compared to manual measurements. A nozzle flow model was designed to determine the spray decision coefficient. When the duty cycle ranged from 30% to 90%, the correlation coefficient of the flow model exceeded 0.95, and the actual flow rate of the system was similar to the theoretical flow rate. Field experiments were conducted to evaluate the spraying effectiveness of the variable spraying control system based on the Kinect sensor. The experimental results indicated that the variable spraying control system demonstrated good consistency between the theoretical spray volume and the actual spray volume. In deposition tests, compared to constant-rate spraying, the droplets under the variable-rate mode based on canopy volume exhibited higher deposition density. Although the amount of droplet deposit and coverage slightly decreased, they still met the requirements for spraying operation quality. Additionally, the variable-rate spray mode achieved the goal of reducing pesticide use, with a maximum pesticide saving rate of 57.14%. This study demonstrates the feasibility of the Kinect sensor in guiding spraying operations and provides a reference for their application in plant protection operations.

Citrus diseases severely impact fruit yield and quality. To facilitate in-field, non-destructive, and rapid detection of citrus leaf physiological and pathological conditions, this study proposes a classification method for citrus leaf physiological and pathological statuses that integrates visible/near-infrared multispectral technology with deep learning. First, a handheld spectrometer was employed to acquire spectral images of five sample categories—Healthy, Huanglongbing, Yellow Vein Disease, Magnesium Deficiency and Manganese Deficiency. Mean spectral data were extracted from regions of interest within the 350–2500 nm wavelength range, and various preprocessing techniques were evaluated. The Standard Normal Variate (SNV) transformation, which demonstrated optimal performance, was selected for data preprocessing. Next, we innovatively introduced an adaptive spectral positional encoding mechanism into the Transformer framework. A lightweight, learnable network dynamically optimizes positional biases, yielding the ASTransformer (Adaptive Spectral Transformer) model, which more effectively captures complex dependencies among spectral features and identifies critical wavelength bands, thereby significantly enhancing the model’s adaptive representation of discriminative bands. Finally, the preprocessed spectra were fed into three deep learning architectures (1D-CNN, 1D-ResNet, and ASTransformer) for comparative evaluation. The results indicate that ASTransformer achieves the best classification performance: an overall accuracy of 97.7%, underscoring its excellent global classification capability; a Macro Average of 97.5%, reflecting balanced performance across categories; a Weighted Average of 97.8%, indicating superior performance in classes with larger sample sizes; an average precision of 97.5%, demonstrating high predictive accuracy; an average recall of 97.7%, showing effective detection of most affected samples; and an average F1-score of 97.6%, confirming a well-balanced trade-off between precision and recall. Furthermore, interpretability analysis via Integrated Gradients quantitatively assesses the contribution of each wavelength to the classification decisions. These findings validate the feasibility of combining a handheld spectrometer with the ASTransformer model for effective citrus leaf physiological and pathological detection, enabling efficient classification and feature visualization, and offer a valuable reference for disease detection of physiological and pathological conditions in other fruit crops.

The rapid and accurate identification of citrus leaf diseases is crucial for the sustainable development of the citrus industry. Because citrus leaf disease samples are small, unevenly distributed, and difficult to collect, we redesigned the generator structure of FastGAN and added small batch standard deviations to the discriminator to produce an enhanced model called FastGAN2, which was used for generating citrus disease and nutritional deficiency (zinc and magnesium deficiency) images. The performance of the existing model degrades significantly when the training and test data exhibit large differences in appearance or originate from different regions. To solve this problem, we propose an EfficientNet-B5 network incorporating adaptive angular margin (Arcface) loss with the adversarial weight perturbation mechanism, and we call it EfficientNet-B5-pro. The FastGAN2 network can be trained using only 50 images. The Fréchet Inception Distance (FID) and Kernel Inception Distance (KID) are improved by 31.8% and 59.86%, respectively, compared to the original FastGAN network; 8000 images were generated using the FastGAN2 network (2000 black star disease, 2000 canker disease, 2000 healthy, 2000 deficiency). Only images generated by the FastGAN2 network were used as the training set to train the ten classification networks. Real images, which were not used to train the FastGAN2 network, were used as the test set. The average accuracy rates of the ten classification networks exceeded 93%. The accuracy, precision, recall, and F1 scores achieved by EfficientNet-B5-pro were 97.04%, 97.32%, 96.96%, and 97.09%, respectively, and they were 2.26%, 1.19%, 1.98%, and 1.86% higher than those of EfficientNet-B5, respectively. The classification network model can be successfully trained using only the images generated by FastGAN2, and EfficientNet-B5-pro has good generalization and robustness. The method used in this study can be an effective tool for citrus disease and nutritional deficiency image classification using a small number of samples.