In response to the dangerous behavior of pedestrians roaming freely on unsupervised train tracks, the real-time detection of pedestrians is urgently required to ensure the safety of trains and people. Aiming to improve the low accuracy of railway pedestrian detection, the high missed-detection rate of target pedestrians, and the poor retention of non-redundant boxes, YOLOv5 is adopted as the baseline to improve the effectiveness of pedestrian detection. First of all, L1 regularization is deployed before the BN layer, and the layers with smaller influence factors are removed through sparse training to achieve the effect of model pruning. In the next moment, the context extraction module is applied to the feature extraction network, and the input features are fully extracted using receptive fields of different sizes. In addition, both the context attention module CxAM and the content attention module CnAM are added to the FPN part to correct the target position deviation in the process of feature extraction so that the accuracy of detection can be improved. What is more, DIoU_NMS is employed to replace NMS as the prediction frame screening algorithm to improve the problem of detection target loss in the case of high target coincidence. Experimental results show that compared with YOLOv5, the AP of our YOLOv5-AC model for pedestrians is 95.14%, the recall is 94.22%, and the counting frame rate is 63.1 FPS. Among them, AP and recall increased by 3.78% and 3.92%, respectively, while the detection speed increased by 57.8%. The experimental results verify that our YOLOv5-AC is an effective and accurate method for pedestrian detection in railways.

The problem of domestic refuse is becoming more and more serious with the use of all kinds of equipment in medical institutions. This matter arouses people’s attention. Traditional artificial waste classification is subjective and cannot be put accurately; moreover, the working environment of sorting is poor and the efficiency is low. Therefore, automated and effective sorting is needed. In view of the current development of deep learning, it can provide a good auxiliary role for classification and realize automatic classification. In this paper, the ResNet-50 convolutional neural network based on the transfer learning method is applied to design the image classifier to obtain the domestic refuse classification with high accuracy. By comparing the method designed in this paper with back propagation neural network and convolutional neural network , it is concluded that the CNN based on transfer learning method applied in this paper with higher accuracy rate and lower false detection rate. Further, under the shortage situation of data samples, the method with transfer learning and ResNet-50 training model is effective to improve the accuracy of image classification.

Metamaterials have become a powerful technique in interdisciplinary research, especially in the field of designing terahertz devices. In this paper, two pairs of different structural units of aluminum–polymer composite metamaterials (APCM) for tunable selectivity are designed. One is designed to achieve high-contrast near-field imaging of linear polarized waves, the other is designed to achieve high-contrast near-field imaging of circularly polarized waves, which means the structural units have very large circular dichroisms. After theoretical design and simulation optimization, it can be found that the contrast of near-field imaging can be effectively controlled by using vanadium oxide (VO2) to fill the open gap of the structure. When the conductivity of VO2 is 200 S/m, both the reflection difference under linear polarization excitation and the reflection difference under the excitation of the circularly polarized wave are at the maximum. The former has a modulation depth of 0.8, and the latter has a modulation depth of 0.55. This work shows excellent tunable selective absorption ability, which will promote the application of metamaterials in terahertz absorber, such as biomedical, non-destructive testing, security inspection, wireless communication and so on.

To meet the demand for high-range and high-frequency impact acceleration measurements, a fiber Bragg grating (FBG) accelerometer with a double-sided symmetrical inclined cantilever beam structure is proposed. The sensing principle is analyzed based on inertial and FBG sensing theories, and theoretical models for sensitivity and resonant frequency are derived. Using PPA-CF material and parameter optimization, the structural dimensions of the cantilever beam, effective fiber length, and proof mass are determined through numerical simulation. Finite element modal analysis shows a first-order natural frequency of 1284 Hz, with the dominant mode being axial translation of the mass block, ensuring pure axial strain on the FBG. A drop-weight impact calibration system is established for experimental testing. Results demonstrate that the sensor achieves a resonant frequency exceeding 1333 Hz, a measurement range above 500 g, and a practical sensitivity of approximately 1.183 pm/g. The linear response exhibits a coefficient of determination, R², of 0.9875, confirming excellent linearity. The proposed accelerometer effectively balances high-frequency response and measurement accuracy, providing a reliable solution for high-frequency impact monitoring in aerospace and impact dynamics applications.