Deep learning techniques have made great progress in the field of target detection in recent years, making it possible to accurately identify plants in complex environments in agricultural fields. This project combines deep learning algorithms with spraying technology to design a machine vision precision real-time targeting spraying system for field scenarios. Firstly, the overall structure scheme of the system consisting of image acquisition and recognition module, electronically controlled spray module and pressure-stabilized pesticide supply module was proposed. After that, based on the target detection model YOLOv5s, the model is lightened and improved by replacing the backbone network and adding an attention mechanism. Based on this, a grille decision control algorithm for solenoid valve group on-off was designed, while common malignant weeds were selected as objects to produce data sets and complete model training. Finally, the deployment of the hardware system and detection model on the electric spray bar sprayer was completed, and field trials were conducted at different speeds. The experimental results show that the improved algorithm reduces the model size to 53.57% of the original model with less impact on mAP accuracy, improves FPS by 18.16%. The accuracy of on-target spraying at 2km/h, 3km/h and 4km/h speeds were 90.80%, 86.20% and 79.61%, respectively, and the spraying hit rate decreased as the operating speed increased. Among the hit rate components, the effective recognition rate was significantly affected by speed, while the relative recognition hit rate was less affected.

Leafy vegetables are an essential source of the various nutrients that people need in their daily lives. The quantification of vegetable phenotypes and yield estimation are prerequisites for the selection of genetic varieties and for the improvement of planting methods. The traditional method is manual measurement, which is time-consuming and cumbersome. Therefore, there is a need for efficient and convenient in situ vegetable phenotype identification methods to provide data support for breeding research and for crop yield monitoring, thereby increasing vegetable yield. In this paper, a novel approach was developed for the in-situ determination of the three-dimensional (3D) phenotype of vegetables by recording video clips using smartphones. First, a smartphone was used to record the vegetable from different angles, and then the key frame containing the crop area in the video was obtained using an algorithm based on the vegetation index and scale-invariant feature transform algorithm (SIFT) matching. After obtaining the key frame, a dense point cloud of the vegetables was reconstructed using the Structure from Motion (SfM) method, and then the segmented point cloud and a point cloud skeleton were obtained using the clustering algorithm. Finally, the plant height, leaf number, leaf length, leaf angle, and other phenotypic parameters were obtained through the point cloud and point cloud skeleton. Comparing the obtained phenotypic parameters to the manual measurement results, the root-mean-square error (RMSE) of the plant height, leaf number, leaf length, and leaf angle were 1.82, 1.57, 2.43, and 4.7, respectively. The measurement accuracy of each indicators is greater than 80%. The results show that the proposed method provides a convenient, fast, and low-cost 3D phenotype measurement pipeline. Compared to other methods based on photogrammetry, this method does not need a labor-intensive image-capturing process and can reconstruct a high-quality point cloud model by directly recording videos of crops.

In view of the problems of high weed regeneration rate in traditional mechanical weeding and environmental risk in chemical weeding, a synergetic strategy of “mechanical damage + wound spraying mechanism” was proposed, and an intelligent weeding device combining synchronous cutting and spraying was designed to enhance the efficacy of herbicides and reduce their use. Focusing on the physical characteristics of weeds and the cutting mechanism, the analysis of the weed-cutting system and the force characteristics of the cutting tool were conducted. Key factors affecting cutting quality were identified, and their respective value ranges were determined. A targeted spraying system was developed, featuring a conical nozzle, DC diaphragm pump, and electromagnetic control valve. The Delta parallel manipulator, equipped with both the cutting tool and nozzle, was designed, and a kinematic model was established for both its forward and inverse movements. Genetic algorithms were applied to optimize structural parameters, aiming to ensure effective coverage of typical weed distribution areas within the working space. A simulated environment measurement was built to verify the motion accuracy of the manipulator. Field experiments demonstrated that the equipment achieved an 81.5% wound weeding rate on malignant weeds in the seedling stage at an operating speed of 0.6 m/s, with a seedling injury rate below 5%. These results validate the high efficiency of the integrated mechanical cutting and targeted spraying system, offering a reliable technical solution for green and intelligent weed control in agriculture. This study fills the blank of only focusing on recognition accuracy or weeding rate under a single weeding method, but lacks a cooperative weeding operation.

Effective management of malignant weeds is critical to soybean growth. This study focuses on addressing the critical challenges of targeted spraying operations for malignant weeds such as Cirsium setosum, which severely threaten soybean yield in soybean fields. Specifically, this research aims to tackle key issues in plant protection operations, including the precise identification of weeds, the lightweight deployment of segmentation models, real-time requirements for spraying operations, and the generalization ability of models in diverse field environments. To address these challenges, this study proposes an improved weed instance segmentation model based on YOLOv8s-Seg, named FCB-YOLOv8s-Seg, for targeted spraying operations in soybean fields. The FCB-YOLOv8s-Seg model incorporates a lightweight backbone network to accelerate computations and reduce model size, with optimized Squeeze-and-Excitation Networks (SENet) and Bidirectional Feature Pyramid Network (BiFPN) modules integrated into the neck network to enhance weed recognition accuracy. Data collected from real soybean field scenes were used for model training and testing. The results of ablation experiments revealed that the FCB-YOLOv8s-Seg model achieved a mean average precision of 95.18% for bounding box prediction and 96.63% for segmentation, marking an increase of 5.08% and 7.43% over the original YOLOv8s-Seg model. While maintaining a balanced model scale, the object detection and segmentation accuracy of this model surpass other existing classic models such as YOLOv5s-Seg, Mask-RCNN, and YOLACT. The detection results in different scenes show that the FCB-YOLOv8s-Seg model performs well in fine-grained feature segmentation in complex scenes. Compared with several existing classical models, the FCB-YOLOv8s-Seg model demonstrates better performance. Additionally, field tests on plots with varying weed densities and operational speeds indicated an average segmentation rate of 91.30%, which is 6.38% higher than the original model. The proposed algorithm shows higher accuracy and performance in practical field instance segmentation tasks and is expected to provide strong technical support for promoting targeted spray operations.

Real-time knowledge of kernel breakage during corn harvesting plays a significant role in the adjustment of operational parameters of corn kernel harvesters. (1) Transfer learning by initializing the DenseNet121 network with pre-trained weights for training and validating a dataset of corn kernels was adopted. Additionally, the feature extraction capability of DenseNet121 was improved by incorporating the attention mechanism of a Convolutional Block Attention Module (CBAM) and a Feature Pyramid Network (FPN) structure. (2) The quality of intact and broken corn kernels and their pixels were found to be coupled, and a linear regression model was established using the least squares method. The results of the test showed that: (1) The MAP^b50 and MAP^m50 of the improved Mask Region-based Convolutional Neural Network (RCNN) model were 97.62% and 98.70%, in comparison to the original Mask Region-based Convolutional Neural Network (RCNN) model, which were improved by 0.34% and 0.37%, respectively; the backbone FLOPs and Params were 3.09 GMac and 9.31 M, and the feature extraction time was 206 ms; compared to the original backbone, these were reduced by 3.87 GMac and 17.32 M, respectively. The training of the obtained prediction weights for the detection of a picture of the corn kernel took 76 ms, so compared to the Mask RCNN model, it was reduced by 375 ms; based on the concept of transfer learning, the improved Mask RCNN model converged twice as quickly with the loss function using pre-training weights than the loss function without pre-training weights during training. (2) The coefficients of determination R² of the two models, when the regression models of the pixels and the quality of intact and broken corn kernels were analyzed, were 0.958 and 0.992, respectively. These findings indicate a strong correlation between the pixel characteristics and the quality of corn kernels. The improved Mask RCNN model was used to segment mask pixels to calculate the corn kernel breakage rate. The verified error between the machine vision and the real breakage rate ranged from −0.72% to 0.65%, and the detection time of the corn kernel breakage rate was only 76 ms, which could meet the requirements for real-time detection. According to the test results, the improved Mask RCNN method had the advantages of a fast detection speed and high accuracy, and can be used as a data basis for adjusting the operation parameters of corn kernel harvesters.

Weed management during the seedling stage is a critical component of soybean production. Efficient weed control can significantly improve crop yield and crop quality. However, conventional spraying techniques exhibit low pesticide utilization and contribute to environmental pollution. To address these challenges, this study proposes a deep learning-based precision target spraying method. A lightweight YOLOv5-MobileNetv3-SE model was developed by replacing the backbone feature extraction network and incorporating an attention mechanism. Field images of weeds were collected to construct a dedicated dataset, and the detection performance of the model was evaluated. Furthermore, a grid-based matching spraying algorithm was developed to synchronize target detection with spray actuation. The system time delay, including image processing delay, communication and control delay, and spray deposition delay, was analyzed and measured, and a time-delay compensation strategy was implemented to ensure accurate spraying. Experimental results demonstrated that the improved model achieved an mAP@0.5 of 86.9%, a model size of 7.5 MB, and a frame rate of 38.17 frames per second. The weed detection accuracy exceeded 92.94%, and spraying accuracy exceeded 85.88% at forward speeds of 1–4 km·h⁻¹. Compared with conventional continuous spraying, the proposed method achieved pesticide reduction rates of 79.0%, 72.5%, 55.8%, and 48.6% at weed coverage rates of 5%, 10%, 15%, and 20%, respectively. The proposed method provides a practical approach for precise herbicide application, effectively reducing chemical usage and minimizing environmental impact.