In order to obtain a band stop filter with filtering function and achieve broadband filtering，this study designed a metal-insulator-metal (MIM) hybrid plasma waveguide，its structure consists of a double sheep horn pin resonator and an MIM straight waveguide. The transmission spectrum characteristics and z-direction magnetic field distribution of this structure were calculated using the finite difference time domain (FDTD) method. The calculation results indicate that a U-shaped region with band stop filtering characteristics is formed between the two Fano-resonance peaks, its filtering band stop width can reach 0.99 m，achieving broadband filtering function. Based on the tunability of the structure, adjusting the distance L, height H, and radius R between the double sheep horn pin resonators can change the resonance wavelength and band stop band. In addition, by adding a new sheep horn pin resonator and forming a cascaded structure, a new Fano-resonance peak can be obtained and the band stop filtering characteristics can be improved. The structural characteristics obtained in this study can provide ideas for the design of optical devices such as band stop filters and sensors.

In order to achieve tunable high-precision filtering and expand communication capacity, this paper presents an innovative filtering system approach that combines dual optical fiber Bragg gratings cascaded with a Sagnac loop. This approach has undergone in-depth analysis and numerical simulation validation using Jones matrix theory. Within the wavelength range from 1 554 nm to 1 570 nm, tunability is achieved by adjusting the length difference between the optical fiber arms to introduce Sagnac phase shifts. This design not only increases the number of filtering channels but also reduces the full width at half height. It is worth noting that under the same arm length difference conditions, this system exhibits a significant improvement in precision compared with single fiber Bragg gratings and Sagnac loop filters. With a power loss of 8 dB, this filtering system allows for an arm length error tolerance of up to 196 nm while providing a more accurate transmission spectrum, making tuning more convenient. This filtering system is easy to implement and widely applicable in dense wavelength division multiplexing systems and the field of optical communication.

Aiming at the problem that the image lacks sufficient clear edge after deblurring by existing methods, a phased image deblurring method based on edge guidance and feature fusion is proposed, and the deblurring task is divided into two stages to gradually remove blur. Firstly, the codec network with double cross integrated attention module (DCIAM) is used to learn the content features of images at different scales to realize the preliminary removal of blur. Secondly, an edge branch network (EBM) is constructed to extract image edge features. Thirdly, an edge-guided deblurring module (EGDM) is designed to couple the content and edge features of images at different resolutions. Finally, cascaded residual blocks and DCIAMs are used to achieve further remove of blur, and a self-calibrated attention fusion module (SCAFM) is introduced to enhance the feature expression. The experimental results demonstrate that the average peak signal-to-noise ratio and structural similarity of the proposed method reach 32.78 dB and 0.964, respectively, which are superior to other comparison methods. The proposed method can significantly improve the deblurring performance and make the image edge structure more complete after deblurring.

To address the problem of mismatch between description statements and image content due to insufficient visual information in image captioning generation, an image captioning generation method based on efficient channel attention network (ECA-Net) is proposed. Firstly, the image segmentation feature as an additional source of visual information, and the iterative independent layer normalization (IILN) module is used to fuse the segmentation feature and grid feature. Also, the image feature is extracted by the double information flow network. Secondly, an ECA-Net module is introduced to the encoder facilitates the learning of correlations among image features through cross-channel interaction, so that the prediction results are more focused on visual content. Finally, the decoder predicts the next phrase based on the provided visual information and the partially generated captions, thus generating accurate captions. Experimental results on MSCOCO data demonstrate that the proposed method can enhance the dependency between the visual information of images, and make the subtitles more relevant and more accurate.

Most high dynamic range imaging methods suffer from uneven brightness, texture loss, and color distortion in the fused images. To address this problem, a high dynamic imaging method based on perceptual priori (PP) and component enhancement is proposed. Based on the perceptual prior, image enhancement is performed on multi-exposure images to retain more texture information and color realism;the luminance component is enhanced by introducing an attention gating mechanism and a feature extraction (FE) operation, and a joint encoder-decoder network enhances the chrominance component. The improved luminance and chrominance components are fused to obtain the final combined image. The experimental data are selected from a sequence of multi-exposure images containing a variety of complex scenes, and the results show that this paper′s method achieves a visual information fidelity of 0.826 8 and a multi-exposure fusion (MEF) structure similarity of 0.947 6. Compared with the fusion images obtained by other methods, the details of this paper′s approach are more precise, the image information is more affluent, it is more in line with the human eye's visual perception, and the fusion effect is better.

In order to make the motion blurred images generated by inspection robots recognize efficiently and accurately during the inspection, a motion blurred crop pest image restoration method based on improved DeblurGAN-v2 was proposed. In order to extract important features of image effectively, the channel attention (CA) mechanism was integrated into the backbone grid of DeblurGAN-v2 to make the model pay more attention to detail features, and improve the restoration ability of motion blurred images. In addition, the spatial pyramid pooling (SPP) was used on the top layer of the original model feature extraction network to alleviate the negative impact of multi-scale changes on image restoration and improve the performance of the model on image restoration. The experimental results of the data set established based on the actual farmland environment show that the PSNR and SSIM indexes of the improved algorithm are 26.281 8 dB and 0.947 3 respectively, which are 8 and 7.2 percentage points higher than the original model. Compared with other mainstream models, the experimental results show that the proposed method has a better effect on the actual restoration of blurred images, and has practical application value to solve the problem of image restoration of crop pests with motion blur.

This paper proposes a new method for defect detection of aluminum profiles to solve the problems of existing methods in accuracy and small target defect detection. Firstly, the coordinate attention mechanism (CA) is introduced into the feature extraction network to prevent information loss, thereby avoiding missed defect detection. Secondly, the adaptive weighted feature pyramid (AWFPN) is introduced to optimize the receptive field of feature maps and attention fusion, thereby improving the efficiency of defect feature capture and utilization. At the same time, the bounding box regression loss function is improved to better handle the scale change and positioning error of defects, thereby improving the detection accuracy and speeding up the convergence speed of the model. Experimental results show that the method has significantly improved the detection effect, especially in the recognition ability of defects such as dirty points and scratches.

Aiming to address the issues of poor stability and low accuracy of laser stripe centerline extraction algorithm in complex environments, a novel centerline extraction method based on improved U2-Net is proposed. Firstly, TSA (transformer-self-attention) and TCA (transformer-cross-attention) modules are added to the U2-Net network to improve the feature extraction ability of the model, achieve accurate pixel-level segmentation, effectively remove noise and glitches in the image, and provide high-quality image sources for subsequent centerline extraction. Secondly, according to the characteristics of the application scenario, the traditional Steger method is improved to complete the high-precision extraction of the centerline of the laser stripe. Finally, the reliability value evaluation mechanism is used to analyze the accuracy of the center point of the light stripe. Experimental results show that compared with other mainstream semantic segmentation networks, the improved U2-Net proposed in this paper has higher extraction accuracy and better anti-noise performance, and the reliability value of the extracted pixel center point on this basis is higher, reaching 1.9 times that of the traditional Steger algorithm, which meets the needs of high-precision industrial measurement.

The principal component analysis network (PCANet), as a network model based on the deep subspace learning framework, has demonstrated remarkable performance in various application domains. However, in the field of rolling bearing fault diagnosis, the PCANet algorithm suffers from issues such as inaccurate reflection of data structural information, poor robustness, and limited generalization ability. To address these issues, this paper proposes a novel rolling bearing fault diagnosis method based on the PCANet algorithm and data augmentation. The proposed method utilizes the L2, 1-norm to learn the frequency domain sparse structure of the rolling bearing vibration signals, effectively suppressing noisy data and enhancing the robustness of the model. Moreover, through the data augmentation processing, the method significantly increases the variability between different classes of the training samples, thereby greatly improving the generalization ability of the model. Finally, experimental results demonstrate that the proposed method significantly enhances the robustness and generalization ability of the PCANet model, enabling accurate identification of different types of the rolling bearing faults.

Free space optical (FSO) communication system using orbital angular momentum (OAM) beams as carriers is highly susceptible to the disturbance of atmospheric turbulence, leading to a degradation of communication quality. This paper proposes a phase compensation scheme based on self-referenced interferometry and analyzes the influence of pinhole aperture and interference intensity ratio on the compensation performance. Compared with classical iterative algorithms via simulations, the results show that the proposed scheme greatly improves computational efficiency while maintaining good compensation performance in weak to medium turbulence and existing slight impairment in strong turbulence.

In order to solve the problem of the high complexity caused by the lack of the early stop strategy in the prefast successive cancellation list (PreFast-SCL) decoding algorithm of polar codes, an improved verification mode of the segmented cyclic redundancy check (CRC) code is proposed, and then combined with the improved verification mode, an enhanced PreFast-SCL (EPreFast-SCL) decoding algorithm is proposed. The proposed algorithm segments the information sequence during encoding and adds the CRC code that can verify the information sequence at the end of each information sequence. In addition, the last CRC code is used to verify the entire information sequence. when a certain segment of the information sequence fails to pass the verification, it is promptly terminated and the error path is eliminated. The simulation results indicate that the proposed EPreFast-SCL decoding algorithm enhances the decoding performance to some extent compared to the cyclic redundancy check aided SCL (CRC-aided SCL, CA-SCL) decoding algorithm and the PreFast-SCL decoding algorithm, with lower decoding complexity.

The finite element simulation software was used to establish the numerical model of laser removal of epoxy resin paint layer on the surface of AH32 steel, and loaded the pulsed laser on its surface for its cleaning process. The effects of different laser parameters on the temperature field of the laser removal of the paint layer on the surface of AH32 marine steel were investigated. The simulation results show that the maximum temperature of the paint layer surface changes with the movement of the spot center, and its scanning trajectory is comet-shaped;the maximum temperature of the paint layer surface decreases with the increase of the spot diameter and scanning speed, and increases with the increase of the laser repetition frequency and the laser power, and verify this law through the in-kind cleaning experiments, and observe the surface morphology of the cleaned surface, and get the results of the cleaning process when the laser power of 28 W, spot diameter of 0.05 mm, repetition frequency of 50 kHz and the scanning speed is 1 000 mm/s, not only can the paint layer be completely removed, but also the substrate material will not be damaged. The results of the study provide theoretical basis and experimental reference for laser cleaning of marine steel surface paint layer.

In order to solve the drawbacks of color reabsorption and ratio tuning among phosphor mixtures, a series of Ba4-xCe3Na3(PO4)6F2:xEu2+ color-tunable luminescent materials are synthesized successfully by high-temperature solid-state method. Under the 297 nm excitation, the emission spectra of the phosphor consist of two emission bands, the yellow light with the peak at 570 nm is assigned to the 4f65d1→4f7 transition of Eu2+, and the luminescence centers red-shifted from 380 nm to 470 nm is assigned to the transition from 4f65d1 to 4f7 of Ce3+. With increasing the Eu2+ doping concentration, the emission region of phosphor gradually moves from the blue region to the yellow region, and the chromaticity coordinates (CIE) also turn from (0.056, 0.069) to (0.303, 0.347). Combined with the fluorescence decay curve, it proves that there is an obvious energy transfer from Ce3+ to Eu2+.

High frequency optoelectronic communication devices require increasingly high piezoelectric material properties, and aluminum nitride (AlN) thin films have attracted attention due to their excellent piezoelectric properties. The deposition process optimization of AlN thin film with high orientation and high-piezoelectric performance has become one of the bottleneck technologies for its application and expansion. Among many preparation process parameters, temperature is one of the key parameters affecting the lattice orientation of AlN thin films. In this paper，aluminum nitride thin films were deposited by magnetron sputtering system，and the effect of deposition temperature on AlN thin film crystal orientation and piezoelectric performance was discussed. X-ray diffraction (XRD) results show that (100) oriented AlN can be obtained under 250 ℃ deposited temperature, and (002) oriented AlN is preferred under 300 ℃;The piezoelectric test results show that piezoelectric constant d33* reaches its maximum value of 0.79 V under 250 ℃, exhibiting excellent piezoelectric characteristics. Based on atomic force microscopy (AFM) and effect of temperature on stability of valence bonds, the mechanism of temperature on its growth was explored.