Search

The RGB image of coal foreign objects lacks target space and edge information, the color and texture between the object to be detected and the background are similar, the contrast is low, and there are overlapping and occlusion phenomena among the objects to be detected, resulting in insufficient feature extraction of coal foreign objects, and the existing foreign object detection methods are difficult to achieve ideal results. In order to solve the above problems, a coal foreign object detection method based on cross modal attention fusion is proposed. By introducing Depth images to construct a dual feature pyramid network (DFPN) for RGB images and Depth images, a shallow feature extraction strategy is adopted to extract low-level features of Depth images. Basic features such as deep edges and deep textures are used to assist deep features of RGB images, effectively obtaining complementary information between the two features. It thereby enriches the spatial and edge information of foreign object features and improves detection precision. A cross modal attention fusion module (CAFM) based on coordinate attention and improved spatial attention is constructed to synergistically optimize and fuse RGB features and Depth features. It enhances the network's attention to the visible parts of occluded foreign objects in the feature map, and improves the precision of occluded foreign object detection. Finally, regional convolutional neural network (R-CNN) is used to output the classification, regression, and segmentation results of coal foreign objects. The experimental results show that in terms of detection precision, the average segmentation precision AP of the proposed method is 3.9% higher than the better Mask transformer in the two-stage model. In terms of detection efficiency, the proposed method has a single frame detection time of 110.5 ms, which can meet the real-time requirements of foreign object detection. The coal foreign object detection method based on cross modal attention fusion can assist color, shape, and texture features with spatial features. It accurately recognizes the differences between coal foreign objects and between coal foreign objects and conveyor belts, effectively improves the detection precision of complex feature foreign objects. It reduces false alarms and missed detections, and achieves precise detection and pixel level segmentation of coal foreign objects under complex features.

Abstract Background Acute respiratory distress syndrome (ARDS) is a serious life threatening clinical critical illness. ARDS-related pulmonary fibrosis is a common complication of ARDS. The occurrence of early pulmonary fibrosis indicates a higher incidence and mortality of multiple organ failure. LPS-induced ARDS-related pulmonary fibrosis model in mice was established in this study. And we have explored the anti-pulmonary fibrosis effects and molecular mechanisms of the Citrus Alkaline Extracts (CAE) in vivo and in vitro. Methods Pulmonary fibrosis mouse model and lung epithelial cell injury model were established in this study. H&E, Masson and Sirius Red staining were used to estimate lung tissue damage. Immunohistochemistry and western blotting were used to analyze proteins expression. Protein-protein interaction was observed by Co-Immunoprecipitation. Systemic impact of CAE on signaling pathway was examined by RNA-seq. Results Through H&E, Masson and Sirius Red staining, it was convincingly indicated that therapeutic administration of CAE alleviated lung injury and fibrosis, while pretreated administration of CAE showed weak improvement. In vitro experiments showed that CAE had dual regulation to E-cadherin and N-cadherin, the important indicators of epithelial-mesenchymal transition (EMT). And it was further demonstrated that CAE reversed TGF-β1-induced EMT mainly through Wnt/β-catenin, Stat3/6 and COX2/PGE2 signals. Through RNA-Seq, we discovered important mechanisms by which CAE exerts its therapeutic effect. And network pharmacology analysis demonstrated core potential targets of CAE in EMT. Conclusion Thus, this study provides new therapeutic effects of CAE in anti-fibrosis, and offers potential mechanisms for CAE in LPS-induced pulmonary fibrosis. Graphical Abstract

Aiming at the problem of poor imaging quality of coal and gangue surface features and low identification rate of coal and gangue caused by insufficient illumination conditions in the process of online coal separation,a new high-quality coal gangue image acquisition that integrates multiple factors such as light source distribution, color temperature, and light intensity is proposed. First of all, in view of the uneven illumination caused by the complexity of the illumination conditions in the actual washing process of coal gangue, the influence of different incident angles on the uniformity of illumination is studied based on the nine-point sampling method, and the best incident angle suitable for coal flow is determined. Then, in view of the problem of coal gangue image distortion caused by different color temperatures caused by different color reproducibility, indicators such as MSE, PSNR, and SSIM are used to quantitatively analyze the image distortion of multi-sample single type coal gangue, and the TOPSIS algorithm is used to study the effect of light sources with different color temperatures on coal and gangue. And the influence of the image distortion of the three different working conditions of mixed coal gangue to determine the color temperature with the best image quality. Finally, considering that online light intensity changes affect the expression of coal gangue surface feature information, based on the response curve of coal gangue surface characteristics under different illuminances, the correlation relationship between exposure time, belt speed and ambient light intensity is established to quantify the illuminance interval with greater surface discrimination.and determine the optimal light conditions. The coal gangue recognition system is established by fusion of high-quality image acquisition methods under multi-factor lighting conditions, and the SSD and Faster-RCNN target detection models are experimentally verified. The results show that this method has greatly improved the image quality of coal gangue, and the accuracy of coal gangue identification has been increased by 10.5%, providing more accurate data support for the intelligent coal gangue sorting system, and has certain application and promotion value for improving the raw coal selection rate.

β-Ecdysone is a phytosteroid derived from multifarious medicinal plants, such as Achyranthes root (Achyranthes bidentata) and Tinospora cordifolia, possessing the potential anti-osteoporosis effect. However, the underlying mechanisms for β-ecdysone treating osteoporosis remain unclear. This study aims to explore the molecular mechanisms of β-ecdysone against osteoporosis by network pharmacology and molecular docking. First, the potential targets of β-ecdysone and osteoporosis were predicted by public databases. Protein interaction and functional enrichment analyses of potential targets were performed using the STRING and Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway databases. Finally, hub targets were identified from network pharmacology, and their interaction with β-ecdysone was validated by molecular docking. Results showed that 47 potential targets were related to the mechanisms of β-ecdysone treating osteoporosis. Enrichment analyses revealed that the potential targets were mainly associated with steroid biosynthetic and metabolic processes, as well as HIF-1 and estrogen signaling pathways. By protein–protein interaction network analysis, top 10 hub targets were screened, including TNF, ALB, SRC, STAT3, MAPK3, ESR1, PPARG, CASP3, TLR4, and NR3C1. Molecular docking showed that β-ecdysone had good affinity with TLR4, TNF, and ESR1. Therefore, β-ecdysone might exert therapeutic effect on osteoporosis development via targeting TLR4, TNF, and ESR1 and regulating HIF-1 and estrogen pathways.

Existing coal and gangue sorting methods mainly use image processing and pattern recognition technology to identify and sort coal and gangue according to texture characteristic values of coal and rock. The grain size of coal and gangue is 25-150 mm, while the coal and gangue above 150 mm still relies on manual sorting. In order to sort coal and gangue with large grain size, a multi-manipulator coal and gangue sorting robot system based on machine vision was proposed.The system uses machine vision to collect coal and gangue information and applies deep learning method to realize coal and gangue identification and grab feature extraction. After obtaining the sequence information of coal and gangue, the sorting work is carried out according to the position of coal and gangue, and the grasping task is assigned to the corresponding manipulator controller by multi-objective task assignment strategy. After the manipulator obtains the task, the target is dynamically monitored according to the acquired task. When the target enters the working space of the manipulator, the visual servo system drives the manipulator to complete the coal and gangue sorting. The test results show that the system can efficiently and quickly sort coal and gangue with grain size of 50-260 mm, and the adopted coal and gangue identification method and sorting strategy have good stability and accuracy under different belt speeds, and the comprehensive accuracy of coal and gangue identification and positioning can reach 93%, which verifies the feasibility of the system.

For complicated structure and control of existing wheeled or crawler-type inspection robot in coal mine tunnel as well as low reliability, and high cost and occupation of tunnel space resource of tracked inspection robot, a kind of suspension line inspection robot in coal mine roadway was designed, which was not influenced by complex geographical environment of roadway and distribution of personnel and equipments. Structure design of the robot was focused on. A virtual prototype model of the robot was established in ADAMS software, and three motion states of the robot were simulated and analyzed including horizontally walking with obstacle navigation, uphill walking with obstacle navigation and downhill walking with obstacle navigation. The results show that the robot always maintains a uniform linear motion in horizontal direction and fluctuates in vertical and horizontal side-sway directions within allowable range, and displacement changes of the robot are relatively stable in the three motion states. The simulation results verify rationality of structure design of the robot.