In the contemporary context, pest detection is progressively moving toward automation and intelligence. However, current pest detection algorithms still face challenges, such as lower accuracy and slower operation speed in detecting small objects. To address this issue, this study presents a crop pest target detection algorithm, GLU-YOLOv8, designed for complex scenes based on an enhanced version of You Only Look Once version 8 (YOLOv8). The algorithm introduces the SCYLLA-IOU (SIOU) loss function, which enhances the model generalization to various pest sizes and shapes by ensuring smoothness and reducing oscillations during training. Additionally, the algorithm incorporates the Convolutional Block Attention Module (CBAM) and Locality Sensitive Kernel (LSK) attention mechanisms to boost the pest target features. A novel Gated Linear Unit CONV (GLU-CONV) is also introduced to enhance the model’s perceptual and generalization capabilities while maintaining performance. Furthermore, GLU-YOLOv8 includes a small-object detection layer with a feature map size of 160 × 160 to extract more features of small-target pests, thereby improving detection accuracy and enabling more precise localization and identification of small-target pests. The study conducted a comparative analysis between the GLU-YOLOv8 model and other models, such as YOLOv8, Faster RCNN, and RetinaNet, to evaluate detection accuracy and precision. In the Scolytidae forestry pest dataset, GLU-YOLOv8 demonstrated an improvement of 8.2% in mAP@0.50 for small-target detection compared to the YOLOv8 model, with a resulting mAP@0.50 score of 97.4%. Specifically, on the IP102 dataset, GLU-YOLOv8 outperforms the YOLOv8 model with a 7.1% increase in mAP@0.50 and a 5% increase in mAP@0.50:0.95, reaching 58.7% for mAP@0.50. These findings highlight the significant enhancement in the accuracy and recognition rate of small-target detection achieved by GLU-YOLOv8, along with its efficient operational performance. This research provides valuable insights for optimizing small-target detection models for various pests and diseases.

According to the characteristics of massive, multi-source, heterogeneous, and rapid growth of book literature data information from the perspective of the metaverse, in order to meet the requirements of efficient management and rapid retrieval such as standardized storage, effective extraction, and scientific library construction for unstructured massive and heterogeneous book in-formation, this study focuses on the normalization of multi-source heterogeneous massive book data, the construction of a warehouse model for book data in the metaverse perspective, and the query and optimization of book data. Systematic research and implementation were conducted to solve the problem of how to process, manage, and query multi-source heterogeneous massive book data in the metaverse, improving the utilization value and query efficiency of the data. This study utilized the semi-structured features of book text data to construct an extraction rule model for heterogeneous book data, and effectively extracted massive heterogeneous book information. Based on the HBase distributed storage structure and parallel computing technology, the storage scheme has been optimized and query efficiency has been improved to ensure efficient management and retrieval of massive heterogeneous book data. The experimental results show that compared with traditional methods, there are significant improvements in multiple aspects such as the accuracy and recall rate of book text data extraction, the management methods and query efficiency of book information.

The three-dimensional reconstruction of forest ecological environment by low-altitude remote sensing photography from Unmanned Aerial Vehicles (UAVs) provides a powerful basis for the fine surveying of forest resources and forest management. A stereo vision system, D-SLAM, is proposed to realize simultaneous localization and dense mapping for UAVs in complex forest ecological environments. The system takes binocular images as input and 3D dense maps as target outputs, while the 3D sparse maps and the camera poses can be obtained. The tracking thread utilizes temporal clue to match sparse map points for zero-drift localization. The relative motion amount and data association between frames are used as constraints for new keyframes selection, and the binocular image spatial clue compensation strategy is proposed to increase the robustness of the algorithm tracking. The dense mapping thread uses Linear Attention Network (LANet) to predict reliable disparity maps in ill-posed regions, which are transformed to depth maps for constructing dense point cloud maps. Evaluations of three datasets, EuRoC, KITTI and Forest, show that the proposed system can run at 30 ordinary frames and 3 keyframes per second with Forest, with a high localization accuracy of several centimeters for Root Mean Squared Absolute Trajectory Error (RMS ATE) on EuRoC and a Relative Root Mean Squared Error (RMSE) with two average values of 0.64 and 0.2 for t_rel and R_rel with KITTI, outperforming most mainstream models in terms of tracking accuracy and robustness. Moreover, the advantage of dense mapping compensates for the shortcomings of sparse mapping in most Smultaneous Localization and Mapping (SLAM) systems and the proposed system meets the requirements of real-time localization and dense mapping in the complex ecological environment of forests.

The effective and comprehensive utilization of forest resources has become the theme of the global “dual-carbon strategy”. Forestry restructured wood is a kind of wood-based panel made of wood-based fiber composite material by high-temperature and high-pressure restructuring–molding, and has become an important material in the field of construction, furniture manufacturing, as well as derivative processing for its excellent physical and mechanical properties, decorative properties, and processing performance. Taking Medium Density Fiberboard (MDF) as the recombinant material as the research object, an event-triggered synergetic control mechanism based on interventional three-way decision making is proposed for the viscoelastic multi-field coupling-distributed agile control of the “fixed thickness section” in the MDF continuous flat-pressing process, where some typical quality control problems of complex plate shape deviations including thickness, slope, depression, and bump tend to occur. Firstly, the idea of constructing the industrial event information of continuous hot pressing based on information granulation is proposed, and the information granulation model of the viscoelastic plate shape process mechanism is established by combining the multi-field coupling effect. Secondly, an FMEA-based cyber granular method for diagnosing and controlling the plate thickness diagnosis and control failure information expression of continuous flat pressing is proposed for the problems of plate thickness control failure and plate thickness deviation defect elimination that are prone to occur in the continuous flat-pressing process. The precise control of the plate thickness in the production process is realized based on event-triggered control to achieve the intelligent identification and processing of the various types of faults. The application test is conducted in the international mainstream production line of a certain type of continuous hot-pressing equipment for the production of 18 mm plate thickness; the synergistic effect is basically synchronized after 3 s, the control accuracy reaches 30%, and the average value of the internal bond strength is 1.40, which ensures the integrity of the slab. Practical tests show that the method in the actual production is feasible and effective, with detection and control accuracy of up to ±0.05 mm, indicating that in the production of E0- and E1-level products, the rate of superior products can reach more than 95%.

With the development of science and technology, it has had a great impact on the manufacturing industry. The manufacturing industry has higher and higher requirements for industrial manipulators, and the expectations for the quality and function of the industrial manipulators in the manufacturing industry are increasing day by day. At the same time, the requirements for the working accuracy of the manipulators are also getting higher and higher. In this paper, a new method is proposed. An error compensation and calibration algorithm based on parameter separation and integration of the standard DH model (SDH) is used to construct the kinematic model of the manipulator. The six groups of structural parameters of the six-axis manipulator are calibrated once every two groups, thereby eliminating the influence of the singular value of the position of the manipulator. Finally, the feasibility of this scheme is verified by simulation experiments, the accuracy is ensured, and the expected purpose is achieved. Because the method in this paper has high requirements on the posture of the manipulator, it reduces the overall calibration calculation and saves time, which has a certain effect on the improvement of the industrial manipulator in the manufacturing industry.