Mobile communication networks are developing rapidly, with a surge in data traffic and intelligent devices. Resource consumption is a huge amount of data, so how to allocate network resources reasonably and efficiently will be the focus of future research. Firstly, a brief overview of the research background and significance of visible light communication (VLC) and radio frequency (RF) heterogeneous network access point (AP) resource allocation is provided; Secondly, regarding the main system architecture of resource allocation in VLC/RF heterogeneous networking, this paper discusses and analyzes the AP allocation technology based on matching and game theory, as well as the research progress in resource management systems; Finally, provide suggestions for the next step of research and provide a summary and outlook.

At present, with the rapid development of optoelectronic chip technology, high-frequency millimeter wave communication has become an important development technology in broadband access networks. Due to the high efficiency and stability of transmitting and receiving millimeter wave signals, the optical millimeter system using an external modulation frequency-doubling scheme can make full use of optical fiber resources in the existing telecommunication network, so it has potential application value in practice. The basic principle of frequency-doubling, typical frequency-doubling simulation or experimental schemes, and related test results are reviewed. Considering that the Mach-Zehnder Modulator (MZM) is a key electro-optical device for frequency-doubling optical generation of millimeter wave, therefore, the key operating parameters are analyzed. The application prospect of optical millimeter waves technology in Radio over Fiber (RoF) systems is also discussed.

Traditional electromagnetic induction current transformers have problems such as large and complex structure, slow response speed, and low safety level, so scholars combine various sensing technologies with optical fibers to explore optical current transformers that can become the main application of power systems in the future. With the mature development of optoelectronic technology, optical current transformers came into being. It has outstanding advantages such as high sensitivity, high integration, high stability and the ability to work in complex environments. In view of the classification of optical current transformers according to different principles, the research status of the principle, structure, manufacturing technology and signal processing of all-fiber current transformers, current transformers based on magnetostrictive effect, current transformers based on magnetostrictive effect, current transformers based on magneto-optical effect, and current transformers based on thermal effect are reviewed, and their future development trends are predicted.

This paper presents the design of a highly stable semiconductor laser diode driver powered by FPGA. The driver is based on the Howland circuit principle, and the FPGA control method is adopted to realize the output current control of 0-500 mA and the current stability of 0.003% when the output current is 300 mA. When driving a 1 550 nm semiconductor laser with a butterfly package, the output optical power reaches 7.49 mW, and the optical power stability reaches 0.33%, which has an important significance for improving the stability of semiconductor laser output optical power.

A miniaturized semiconductor laser driver and temperature control circuit based on STM32F407 microcontroller combined with FPGA is designed to address the issue of the wavelength and power output of tunable semiconductor lasers being unable to remain constant due to fluctuations in their operating current and temperature. By interacting with the host computer, the system can achieve continuous adjustment of the laser driving current and operating temperature, ensuring that the wavelength of the output light of the semiconductor laser is within the optimal range. After testing, the current stability of the drive module of the system is 0.075%, and the average relative error between the actual and theoretical values of the current is 0.538%; The stability of the temperature circuit is 0.03 ℃, fully meeting the practical application requirements.

A stable control technique based on Photoelastic Modulation (PEM) for resonant phase tracking of driving signal is studied. The equivalent circuit model of PEM and its resonance matching circuit is established, and the amplitude-frequency and phase-frequency relationships of driving voltage and inductance voltage of the matching circuit are deduced, as well as the relationship between PEM resonance frequency and inductance voltage resonance phase. A PEM control system is set up, with the feedback voltage phase of the initial resonance frequency as the zero phase reference. Adjusting the drive voltage stabilizes the feedback voltage phase near the phase reference value, and corrects the phase reference value with the change of the resonance frequency so that the drive voltage frequency is always stable near the PEM resonance frequency. Taking the phase delay amplitude of 1/4 wave and half wave PEM as examples, the stability control tests for 60 min are performed. The results show that: (1) When PEM works in 1/4 wave state, q=0.001 4 rad, the maximum relative deviation is 0.58%; (2) Standard deviation when working in half-wave state, h=0.001 3 rad with a maximum relative deviation of 0.32%.

To develop a high repetition rate nanosecond narrow-pulse light source with with pulse width, repetition frequency and output power tunable to the single-photon level for the visible wavelength with high repetition frequency, a green frequency-doubling light source design based on a gain-modulated DFB laser and a PPLN crystal is proposed. The DFB laser is driven by a homemade nanosecond high-speed driver circuit in a pulsed mode to generate a fixed-wavelength pump source; using a fiber optic semiconductor attenuator to adjust the attenuation amount and modulate the intensity of pump light; using PPLN crystal as a nonlinear crystal material, a quasi phase matched green frequency doubling laser output was obtained through its tunable properties. The experimental results show that the pump power and center wavelength of the gain modulated DFB laser are stable, with a pulse width of 1.02~10.25 ns and an operating frequency of 1 Hz~10 MHz that can be continuously and independently adjusted. When the power of pump light decays to -44.61 dB, the average photon number of the green frequency doubling light output pulse is about 384N, and the conversion efficiency of the frequency doubling light to light is 9.3%.

The research progress of optical hydrogen sensing technology based on surface plasmon (SP) is summarized, and two measurement methods based on SP, namely localized surface plasmon resonance (LSPR) and surface plasmon resonance (SPR), and then the measurement method, sensing structure and sensing characteristics of the sensor based on these two principles are expounded. In addition, various existing optical hydrogen sensors are reviewed, and their advantages and disadvantages are introduced. Finally, we emphasize the development direction and important challenges of SP-based optical hydrogen sensors in the future, such as special materials with specificity and stability, new nanostructures that are constantly developed and applied, and multi-point sensors that can meet future needs.

In order to realize the monitoring of crop growth and combine with the development trend of fluorescent lidar technology, a set of fluorescent lidar system suitable for crop monitoring is built by using laser induced fluorescence technology in this paper. In order to improve the stability of the system and simplify the structure of the system, the laser transmitting system and the receiving optical system are combined into one. By adjusting the negative lens to the position of the cassegrain system, the laser transmitting and signal receiving of distant objects can be carried out. At the same time, in order to improve the stability of the system, a coaxial correction method and device of the lidar system are proposed. By adding a two-dimensional adjustable stop matching the field of view to the receiving optical system, the focal length adjustment problem is realized in the scene of distance change, and the detection ability of the received target signal is enhanced. In this study, the whole fluorescence lidar system was constructed and coaxial installation was carried out, and the fluorescence spectra of the blades were measured and verified. The main fluorescence characteristic peaks of the blades were detected at 80 m,100 m and 150 m respectively. The experiment proves that the fluorescence signal obtained by coaxial adjustment is of good quality and the system is stable. After the crop distance changes, the high SNR fluorescence measurement of samples at different distances can be directly realized by adjusting the system, which provides theoretical and technical support for the fluorescence lidar system in agricultural crop detection.

Aiming at the problems such as high complexity and high labeling cost of the traditional directed target detection algorithm based on rotating frame labeling, a weakly supervised directed target detection algorithm LSK-EFPN based on cross-space and multi-scale is proposed, which can infer the rotating frame information of the target by using the horizontal frame labeling information. The directed target detection in complex remote sensing scenarios is realized. In order to improve the network detection ability, the algorithm uses LSKNet network to extract the prior background features of input images, and adds a cross-space multi-scale attention module to capture cross-space feature regions. Finally, CIoU is used as a scale-constrained loss function to reconstruct the consistency loss. The experimental results show that the average accuracy of LSK-EFPN on DIOR data set of remote sensing scenes reaches 61.7%, which is 4.7% higher than H2RBox algorithm, providing a new technical solution for directed target detection scenes based on horizontal box marking.

The tension clamp plays the role of connecting wires and carrying current in the power line, and its crimping quality is directly related to the safety of the power grid. In order to solve the problems of complex operation and high personnel requirements in the DR defect detection of tension clamp crimping, an EW+YOLOv8 application EW+YOLOv8 tension clamp DR image defect detection and identification scheme was proposed, which selected YOLOv8n with good accuracy and good real-time performance as the detection reference network, and then added the efficient channel attention mechanism ECA to highlight the key information of the defect feature map, and applied the Wise-IoU loss function based on the dynamic non-monotonic focusing mechanism to replace the CIoU loss function. Reduce the impact of low-quality anchor frames in labeled samples. Based on the data set preparation and the analysis of test evaluation indexes, relevant ablation experiments and comparative experiments were carried out, which showed that EW+YOLOv8n had fast calculation speed and few model parameters when ensuring high detection accuracy, and was applied to the detection and identification of power tension clamp crimping defects with mAP@0.5 and FPS of 97.4% and 50 sheets, respectively, which could meet the actual detection needs of the project.

In pulse coherent lidar wind measurement, the widely used FFT algorithm is simple and fast, but the range resolution of wind measurement is difficult to further improve. Time-frequency analysis methods such as continuous wavelet transform (CWT) have fine time-frequency analysis capabilities, but the calculation The real-time performance is poor. This paper proposes a fusion algorithm that combines the advantages of FFT and fast continuous wavelet transform (fCWT). The algorithm inherits the fine analysis capabilities of CWT, and the computing speed is significantly accelerated. By comparing the radar wind measurement simulation data and actual measured data with the fusion algorithm, CWT, and FFT algorithms, the wind speed curves drawn by the fusion algorithm and the CWT and FFT algorithms have the same trend, but have richer details, and the calculation time of the fusion algorithm is reduced compared to the CWT algorithm. more than 45%. This fusion algorithm provides a new idea for improving the wind measurement performance of pulse coherent laser wind measurement radar.

In response to the problem of low signal-to-noise ratio and frequency dependent detection accuracy in the detection of vehicle vibration signals using a phase sensitive time-domain reflectometer (Phase Sensitive Optical Time Domain Reflectometer, -OTDR) system. A Guided Edge Filter (GEF) method has been proposed. The GEF method preserves and extracts effective vibration information while reducing the impact of noise. Use PZT to generate vibration signals of different frequencies. Compare the GEF method with the direct amplitude difference method experimentally. The results show that the signal-to-noise ratio obtained using the GEF method has an average improvement of 3.26 dB compared to the direct amplitude difference method. Using the vibration signals of vehicles driving on the road for testing, the results show that the highest signal-to-noise ratio obtained using the GEF method is 3.98 dB, which is 2.65 dB higher than the direct amplitude difference method.

Aiming at the problem of misdetection and omission caused by dense vehicle targets, mutual occlusion and too small targets in automatic driving scenarios, a vehicle target detection algorithm with improved YOLOv7 is proposed. The ACmix hybrid attention mechanism is added after the SPPCSPC of the backbone network to fully mine the feature information, enhance the network's attention to the vehicle information, reduce the interference of other targets, and improve the detection accuracy; the Swin Transformer is added to the Neck end to collect the global information; the 160×160-size target detection head is added to increase the number and density of the anchors and to improve the network's ability to perceive small targets; finally, Soft-NMS flexible non-maximum suppression is utilized to reject redundant candidate frames and improve the leakage detection ability. The feasibility of the improvement is verified by experiments and compared with five mainstream networks, and the average accuracy reaches 91.5%, and compared with the basic network YOLOv7, the average accuracy is improved by 7.1%, and the operation speed reaches 105 FPS, which proves the effectiveness of the improved method.

In the displacement measurement of fiber grating interference sensor, the measurement results of the sensor will fluctuate or have errors due to the influence of environmental confounding such as external vibration or electromagnetic. In order to improve the efficiency of displacement measurement using fiber Bragg grating interferometric sensors and ensure the accuracy of measurement, a high-precision calibration method for displacement measurement error of fiber Bragg grating interferometric sensors is proposed. Obtain displacement measurement data by analyzing the displacement measurement principle of fiber Bragg grating interferometric sensors; Using wavelet threshold method to denoise and smooth displacement measurement data; Using Back Propagation (BP) neural network, high-precision calibration of error data in displacement measurement data after denoising is carried out. The experimental results show that the proposed method can effectively improve the accuracy, efficiency, and operational stability of error calibration while reducing the computational complexity of error calibration.

Image fog removal is an important research focus in image processing field. In order to solve the problem of image defogging and enhancement in hazy weather, a defogging algorithm based on improved dark channel is proposed. First of all, in order to make the haze image closer to the non-fog image and improve the clarity of the image, the algorithm reduces the RGB channel values of the fog image respectively, and combines each reduced channel and the other two previously unreduced channels, and then uses the image de-fog algorithm to weight three new images to restore the image. In order to solve the problem of color distortion in the sky region of the image, a parameter K is set to calculate the transmittance of the sky region and the non-sky region respectively. In order to solve the problem of over-dark brightness and increase target contrast, this paper introduces CLAHE method to enhance image processing. The experimental results show that: The contrast value of the proposed algorithm in the five images is more than twice and more than three times that of the MDCP and DCP algorithms respectively, and the average information entropy in the five images is 7.558 9, which is obviously better than the other two algorithms. Moreover, the average accuracy of the proposed algorithm in target detection under haze weather can reach 73%, which is 7% higher than before the improvement, and has certain feasibility.

Military aircraft target detection in remote sensing images is of great significance for reconnaissance and early warning and intelligence analysis. In view of the challenges of complex image background, large target scale variation and dense distribution in this task, a lightweight detection model based on Faster R-CNN is proposed. The model uses residual split attention network to capture the global context information of target region features to improve the representation ability of the model; it uses deformable convolution to dynamically learn the deformation features of target region, adapt to targets of different scales and shapes; it uses the method of comparative experiment to streamline the backbone network, reduce the impact of too deep backbone network and too low sampling rate on small target detection, and improve the recognition speed of the model. In the target candidate box screening stage, Soft NMS algorithm is introduced to remove candidate boxes with high overlap according to the descending order of confidence, and reduce the miss detection rate of densely distributed targets. The experimental results show that the Faster R-CNN model proposed in this paper has a mAP0.5-0.95 of 77.1% when the number of parameters is 23.844 MB, and the detection speed reaches 43.7 frames per second. Compared with multiple mainstream models, it has better comprehensive performance.

A lightweight YOLOX-S detection model is proposed for industrial production to address the issues of large model size and unsatisfactory detection performance in surface defect detection of solar cells. Firstly, based on the YOLOX-S model, the lightweight network MobileNetV3 is used to optimize the backbone network, reduce model parameters, reduce model computation, and improve detection speed. Secondly, MobileNetV3 is improved by using the FReLU Activation function to make the model have the spatial pixel level modeling ability, improve the sensitivity of model spatial feature information, and enhance the feature extraction ability of the model for small target defects. Finally, an attention feature fusion module is introduced into the neck network to aggregate multi-scale information and enhance the model's multi-scale feature fusion capability. The experimental results show that the average accuracy of the improved YOLOX-S detection model can reach 97.6%, the number of parameters can be reduced by 43.2%, the detection speed can reach 51 frames/s, and the confidence level is above 90%. The detection results are reliable.

Digital hologram provides users with more detailed 3D information; However, its encoding distortion and the unique properties pose challenges to its quality assessment. A full reference hologram quality assessment method combining holographic plane and object plane is proposed. Firstly, the degradation degree of compressed hologram is measured using information fidelity criteria on the holographic plane. Secondly, significance weighted gradient similarity is extracted from the reconstructed multi view reference hologram and compressed hologram on the object plane, and global feature changes are measured using Grnwald-Letnikov fractional derivatives; then the similarity between the reconstructed multi view reference hologram and compressed hologram is calculated based on these two features. Finally, the calculation results of the holographic plane and the similarity of the object plane are used as joint features, and random forest is used to train the above features to establish a quality assessment model. The experimental results show that the Pearson linear correlation coefficient is about 0.917 9 and the Spearman rank correlation coefficient is about 0.907 5, which outperforms the relevant representative hologram quality assessment methods and has a higher degree of fit with the subjective perception of the human eye.

In view of the problems of local Receptive field limitation and deep semantic feature loss when Segformer processes remote sensing images with complex spatial and spectral characteristics, a multi-level layered encoder network structure with different attention modules embedded between Segformer modules at different levels is proposed: polarization attention module PSA is embedded before Block2, To enhance the network's spatial perception of largescale features, alleviate feature semantic loss, and embed efficient channel attention module ECA before Block3 and Block4 to obtain weighted features of the channel, thereby enhancing the network's recognition and perception ability of important features, and ultimately achieving pixel level semantic segmentation of remote sensing images through multiple feature cascades. Through testing on the GID and BCDD datasets, compared to the original Segformer, the new network has increased mIOU (%) by 1.85% and 1.63%, respectively, in both datasets.

In image fusion, most edge-preserving filters corrupt structure and texture information during the optimization process, and the noise will also seriously affect the fusion result, which may cause problems such as loss of details and textures in the fusion results. An infrared and visible image fusion method based on RPCA(Robust principal component analysis) algorithm is proposed in this paper, which can effectively improve figure definition and visual information fidelity. Firstly, infrared and visible light images were decomposed into low rank and sparse images through Robust principal component analysis. Then, relative total variation (RTV) and average energy method were adopted to process low rank and sparse images. Finally, the final fusion image was obtained by inverse NSCT transformation. The experimental results show that, compared with the other methods, the fusion image generated by the method proposed in this paper has certain improvements in the average gradient, spatial frequency, edge intensity and mutual information, with an increase of 10.6% to 72.6%, 15% to 60.2%, 9.7% to 69.6%, and 22.7% to 229.7%, respectively.

The geographic location features of optical remote sensing images are complex and diverse, and the spatial information of multi-scale features is rich, which makes it difficult to fully extract the image features in the registration process and the registration accuracy is low. To address the above problems, an registration model that fuses contextual features and densenet is proposed, which deepens the attention to location information by embedding it in attention and integrates several different kernels of depth-separable convolution to integrate several different sensory fields to aggregate the rich multi-scale feature semantic information. Firstly, the fused densenet is used to extract the feature information from the image, then the bidirectional matching relationship is obtained using bidirectional Pearson correlation matching, and the final parameters are synthesized by weighting the bidirectional parameters obtained through regression, and finally the image registration is completed by affine transformation. The experimental results show that the proportional index coefficients of key points correctly estimated at 0.05, 0.03 and 0.01 are as high as 83.9%, 60.3% and 15.3% in the Aerial-image dataset, respectively, which effectively improves the optical remote sensing image registration accuracy.

This paper proposes an infrared and visible light fusion method based on gradient residual dense blocks and shuffle attention mechanism to address the problem of insufficient extraction of fine-grained details and difficult utilization of deep features in deep learning-based infrared and visible light image fusion. This method incorporates gradient residual dense blocks and attention shuffle modules into the encoder, which enhances the ability of the autoencoder to extract fine-grained details and deep global features and suppress noise. In comparison experiments with other methods, our method shows good performance in subjective evaluation in terms of detail texture and global level, and it effectively fuses the effective features of infrared and visible light source images. In objective evaluation, our algorithm achieves optimal values in five metrics including standard deviation, peak signal-to-noise ratio, visual information fidelity, QAB/F, and wavelet feature mutual information, which are 76.927 5, 16.775 5, 0.876 7, 0.514 1, and 0.431 3.

To improve the accuracy of feature point localization in laser remote sensing images, a remote sensing laser image feature localization technology based on deep learning algorithm is proposed. Using hyperspectral model parameter fusion and spectral band image detection methods, multi frequency band detection and sparse representation are performed on remote sensing laser images, extracting multi-spectral and point cloud data, and recombining features from different information dimensions; Using deep learning algorithms, dynamically iteratively search for physical and geometric feature points of the target. Using image segmentation technology to cluster and process composite features of feature points; Using undirected graphs to model neighborhood relationships for feature localization, accurately locate feature points in remote sensing images based on feature clustering and neighborhood output expression results. The simulation results show that using this method for remote sensing laser image feature localization has a high resolution level and can achieve anti blurring target feature point localization, with a maximum positioning accuracy of 0.92. Under the interference of point cloud noise, the maximum offset is 6 * 10-3.

In order to improve the virtual reconstruction effect of laser 3D images, a laser 3D image virtual reconstruction method based on visual communication technology is proposed. Obtain texture feature values of laser 3D images, collect laser 3D images, and extract laser 3D image reconstruction features; Fusing frequency domain features and texture features to obtain a laser 3D image multi feature dataset and bad features within the laser 3D image multi feature dataset, and designing steps for multi feature fusion; Based on the results of multi feature fusion, a three-dimensional specific laser 3D image model is obtained to achieve virtual reconstruction of laser 3D images. The experimental results show that the system can effectively solve the problem of laser 3D image reconstruction. When the occlusion ratio is 18%, the scale invariant depth error value and average angle error value both reach the maximum, and the maximum value does not exceed the set threshold. The structural similarity of the method in this paper is over 90%, and the reconstruction accuracy is high, which can achieve high-precision virtual reconstruction of laser 3D images.

The parking space enhancement algorithm is an important part of automatic parking, and its enhancement result directly affects the extraction effect of the parking space line. Based on this, this paper introduces the dark channel as a low-frequency component for adaptive contrast enhancement. Based on multiple sets of low-contrast parking space image data, the applicability of various low-contrast enhancement algorithms is discussed. The exposure phenomenon leads to the problem of reducing the integrity and accuracy of parking space line extraction. A fast enhancement algorithm for peak histogram equalization is proposed, which combines PSNR, structural similarity, average brightness and information entropy as objective evaluation indicators, and uses Hough The linear detection statistical algorithm enhances the accuracy of the parking space extraction results and is verified. The research results show that the algorithm in this paper can reduce the interference of environmental information, retain more texture details, improve the brightness and contrast of the global image, and it still has excellent robustness in low-light environments. The algorithm in this paper has a parking space line extraction accuracy of more than 90%, and the algorithm running time is only 37.18 ms, which can provide method guidance for automatic parking systems in low-contrast scenes.

With the booming development of point cloud-based applications such as intelligent driving and robot navigation, semantic segmentation of point clouds has gradually become a hotspot of research. However, the existing methods for semantic segmentation of point clouds suffer from the shortcomings of insufficient local feature extraction and incomplete feature fusion. To address these shortcomings, we propose corresponding solutions. For the phenomenon of insufficient local feature extraction, the explicit features of neighboring points are associated by embedding the coordinates, directions, distances and other relevant information of the neighboring points. For the phenomenon of incomplete feature fusion, a hybrid pooling method combining maximum pooling and self-attention pooling is proposed. The network architecture in this paper is based on PointNet++ and is combined with the proposed local feature extraction and fusion method. The experimental results on the S3DIS dataset show that the evaluation indices have been improved to different degrees compared with baseline PointNet++ method, which confirms the effectiveness and superiority of new method.

Optical signal-to-noise ratio (OSNR) is closely related to the transmission performance of optical fiber communication, so OSNR monitoring is a crucial part of optical performance monitoring technology. At the same time, the dispersion in the transmission link will lead to the broadening of optical signal pulses, which will reduce the accuracy of OSNR monitoring. Aiming at this problem, a convolutional neural network model is designed. The asynchronous delay sampling graph (ADTP) is used as the network input feature, and the instance batch standardization module is introduced. It inherits the advantages of feature divergence distribution at different depths of the neural network and improves the adaptability of the neural network to dispersion changes. The experimental results show that the mean absolute error (MAE) of the model is 0.2 dB in the case of 10 Gb/s NRZ-OOK signal without dispersion interference monitoring, and the MAE is reduced by 0.61 dB at most in the case of dispersion change.

To improve the coding efficiency and decoding correctness of vortex optical communication. In this paper, two vortex light beams carrying different low orbit Angular momentum and radial index are used to stack to produce 16 different light intensity distribution maps, which are encoded with 4-bit binary. To address the impact of atmospheric turbulence on light intensity distribution, a Vision Transformer neural network model optimized by sparse attention algorithm is proposed, and the light intensity distribution map affected by strong turbulence is used as input for training, Thus achieving accurate identification of distorted information. The simulation experiment shows that the accuracy of this model in identifying vortex beams affected by strong turbulence can reach 95.5% and it is more accurate in resolving local details. The model excelled in recognizing accuracy despite strong turbulence, showcasing its robustness and universality across wavelengths and distances.

In order to improve the secure transmission capability of data in fiber optic sensing networks, a secure data import method based on information leakage encryption transmission is proposed for fiber optic sensing networks. The discrete chaotic time series synchronous modulation method is used to realize the modulation and coding processing in the process of importing the transmission data of the optical fiber sensor network. According to the chaotic random coding characteristics, the synchronous output stability adjustment and autonomous random coding in the process of transmitting data of the optical fiber sensor network are realized, and the encryption key for suppressing the information transmission leakage of the optical fiber sensor network is constructed, The information leakage encryption retransmission during the data import process of the optical fiber sensor network transmission is realized through Logistic mapping. According to the mixed sensitive key representation and Arithmetic coding of the optical fiber sensor network transmission data, the information leakage control and data security import are realized. The simulation test results show that the encryption performance of the imported data transmitted by the proposed method is good. The recognition rate of the encrypted bit sequence of the transmitted data transmitted by the optical fiber sensor network is not less than 98%, the highest bit error rate is only 2.7*10-9%, the anti-leakage ability reaches 0.970, and the integrity of the imported data reaches 0.996. It shows that the method has strong encryption effect and anti-leakage ability, and realizes the safe import and encrypted transmission of data.

When laser communication network transmit information, it is easy to be affected by discrete variables, which reduces the quality of information transmission. Therefore, the optimization method of laser communication network topology under the influence of discrete variables is studied. The initial topological structure of laser communication network is generated, and the influence of discrete variables is analyzed. According to the optimization target of laser communication network topology, the optimization target model of laser communication network topology is established. Multi-objective evolution method is adopted to eliminate redundant links in the topological structure and optimize the topological structure of laser communication network. The experimental results show that the average link length of the optimized network topology is less than 50cm, the energy of communication nodes is more than 48J, and the number of network survival rounds is more than 371. The optimization performance is good, and the redundant paths in the network topology are effectively eliminated.

In order to improve the application effect of fiber grating, a high-speed and accurate fiber grating signal demodulation scheme is designed. According to the demodulation principle of LPFG fiber grating signal, two LPFG double-edge filters are designed to process the two reflected light, and the photodetector is used to detect the optical power result, so as to demodulate the central wavelength of the grating. The optical fiber interferometer is used to detect the demodulation frequency and perform positive and negative scanning. The demodulation result is compensated according to the correlation between the demodulation error caused by optical propagation delay and the scanning direction of demodulation frequency. The test results show that the scheme has good feasibility. The standard deviation of demodulation results of FBG signals is less than 1.85pm when the spectrum overlaps in different degrees. The demodulation error of grating wavelength is between -5~5pm. Effectively reduce the demodulation error, improve the demodulation accuracy of fiber grating, and obtain the information in the signal reliably.

The resource utilization of passive optical communication network is directly related to the communication quality of the network. In this context, in order to achieve high-quality communication, this paper proposes a passive optical communication network reconfiguration method based on genetic ant colony algorithm. Under 8 assumptions, the research aims to maximize resource utilization. Under 7 constraints, the genetic ant colony algorithm is used to obtain the optimal solution that meets the objective function, and the passive optical communication network reconfiguration scheme is obtained. The results show that the resource utilization ratio of the passive optical communication network under the application of the design method is high, up to 96%. It shows that the design method can achieve communication transmission with the minimum resource consumption, and has certain application value.

The current vehicle anti-collision control system has shortcomings, such as the inability to identify various obstacles. In order to solve the shortcomings of the current vehicle anti-collision control system and improve vehicle driving safety, a vehicle anti-collision control system based on near-infrared laser radar is designed. Firstly, the current research progress of vehicle anti-collision control system is analyzed to find out the factors causing the deficiency of the system. Then, the overall architecture of the vehicle anti-collision control system is designed. After collecting the data in front of the vehicle by using near-infrared Lidar, the data is pre-processed to avoid obstacles in front of the vehicle. The test results show that the system can effectively overcome the shortcomings of the current vehicle collision prevention control system, collecting data in front of the vehicle with higher accuracy and speed, identify obstacles with high accuracy and sensitivity, and has high practical application value.

In order to obtain the optimal process parameters for 17-4PH precipitation hardened stainless steel powder in the broadband laser cladding process, this study takes laser power, preheating temperature, and scanning speed as controllable input parameters, and takes the aspect ratio, saturation, and wetting angle of the cladding layer as output. Three factors and five levels orthogonal experiments are conducted. The response value nonlinear mathematical model constructed by the Mcquart algorithm is used to explore the influence of optimization variables and their weight ranking on the macroscopic morphology of the coating. The second generation non dominated genetic algorithm is used for optimization, and the output has the Pareto optimal frontier to determine the optimal process parameters. The results show that the optimal combination of process parameters after optimization is laser power of 1 878 W, scanning speed of 14 mm/s, preheating temperature of 200 ℃. At this time, the width to height ratio of the obtained cladding layer is 32.08, saturation is 0.76, and wetting angle is 10.2°. The average error of the optimized output is less than 5%.

In order to improve the firing accuracy of weapons, a baseline detection method of infrared Telescopic sight with laser scattering characteristics is designed. By fixing the Telescopic sight and using the blackbody to emit infrared laser, an infrared parallel beam is formed after the mirror reflection, and the target image is generated in the field of view of the Telescopic sight. The baseline detection is realized by adjusting the Telescopic sight graduation center to coincide with the target image center. The Bidirectional reflectance distribution function and Kirchhoff approximation theory are used to analyze the scattering characteristics of infrared laser and the influence of wavelength and scattering angle on imaging accuracy, so as to improve the accuracy of baseline detection. Experimental proof: When the incident wavelength is 9.5 At around m, the roughness of the shooting target is 8 At m, the bistatic polarization scattering coefficient is the lowest, around 1.0; When the scattering angle is around 29°, the roughness of the shooting target is 8 At m, the lowest bistatic polarization scattering coefficient is around 0.146; The average deviation of the target position after correction is 0.01 m.

The application scope of 3D laser scanning technology is gradually expanding, but the lack of real texture information in laser point clouds results in deviation between target detail reconstruction and actual targets. 3D visualization digital imaging technology can effectively compensate for the shortcomings of 3D laser scanning technology. How to effectively integrate and apply it has become a key research direction, Therefore, a study on the 3D visualization digital image fusion target detail reconstruction method of GeoSLAM laser point cloud is proposed. Using the point cloud filtering algorithm to remove noise information from the GeoSLAM laser point cloud, and correcting and processing the distortion phenomenon of 3D visualization digital images, based on this, the Moravec operator is used to extract the detailed feature points of the fusion target, register the feature points in the laser point cloud and digital images, and develop the fusion target detail reconstruction program, thus achieving effective fusion of the laser point cloud and digital images, that is, precise reconstruction of the target details. The experimental results show that: Compared with the comparison methods, the proposed method can complete the detail reconstruction more accurately. It has an average correlation coefficient of 0.929 5, and a lower error rate of detail feature point extraction of the fusion target. The detail reconstruction effect of the digital image of the architectural category is better, and the error rate of detail feature point extraction is only 1%, which fully confirms the better application performance of the proposed method.

The acquisition of weak magnetic signals is affected by factors such as environmental interference and unreasonable deployment of sensing equipment nodes, resulting in a large amount of interference noise in the collected signals. In order to improve the resolution of weak magnetic signal acquisition, an intelligent acquisition method for weak magnetic signals based on pyroelectric infrared sensors is proposed by combining array signal processing technology. Construct a node distribution array model for the intelligent acquisition of weak magnetic signals using pyroelectric infrared sensors, and combine phase shift correction and noise interference suppression methods to construct a filtering model for weak magnetic signal acquisition. The coherent array detection method based on signal waveform enhances the spatial features of weak magnetic signals, and combines multipath signal coherent matching technology and corresponding multiple DOA estimation techniques, Intelligent signal acquisition is achieved by enhancing and filtering the output signals of the array nodes of pyroelectric infrared sensors. The experimental results show that the output signal amplitude and phase gain of intelligent acquisition of weak magnetic signals using this method are good, with an output signal-to-noise ratio of up to 45.9 dB and a bit error rate of up to 0.55 bit for weak magnetic signals, indicating the practicality of this method.

This article discusses the innovative collaborative application of Light-Controlled Lens (LCL) and Reflector Technology in enhancing the quality of urban street lighting, offering a more efficient and sustainable new solution for main thoroughfare illumination. The combination of LCL and reflectors not only improves lighting efficiency and quality but also significantly reduces energy waste and light pollution. The article also emphasizes the multifaceted collaborative modes of these two technologies, which can adjust lighting based on real-time traffic and environmental conditions, enhancing safety, reducing energy costs, and beautifying urban spaces. This study provides a new perspective for the development of future urban lighting systems, showcasing the great potential of the integration of LCL and Reflector Technology, with outstanding performance in environmental sustainability and adaptation to the needs of smart cities.

The anti-counterfeiting recognition method of product packaging adopts wavelet packet transform algorithm to recognize the anti-counterfeiting information, but it is easy to be affected by the positive and negative characteristics of the anti-counterfeiting image during the recognition, resulting in a low recognition rate. Therefore, the invisible laser holographic anti-counterfeiting identification method of product packaging was proposed. Spatial filtering is used to sharpen the edge of security image and enhance the features of security image. Based on the image correction results, the positive and negative characteristics of the image were analyzed through the real part output and the simulated output of the segmentation filter, and the features of the image were extracted. According to the similarity criterion, the template matching algorithm was used to realize the anti-counterfeiting recognition. The experimental results show that the proposed method can obtain more than 90% recognition rate and the signal-to-noise ratio is above 60 dB when applied to low-quality anti-counterfeiting information.

A dense matrix equalization method for solar laser pulse amplitude is proposed to address the issues of relatively high equalization complexity and high laser power consumption in the current process of solar laser pulse amplitude equalization. Using wavelet transform technology to eliminate noise signals in the pulse amplitude signals of solar lasers; By using signal resampling technology, the original signal is transformed into a signal with a fixed order ratio, and different order components are extracted to grasp the fluctuation characteristics of the pulse amplitude signal; The fluctuation characteristics of the pulse amplitude signal are used as network inputs to achieve pulse amplitude equalization of solar lasers through training. The experimental results show that this method can significantly reduce the complexity and delay of equalization processing, while also reducing the power consumption during the operation of solar lasers.

In order to improve the intelligence and visualization level of high voltage insulator defect intelligent detection, an intelligent detection method of high voltage insulator defect based on infrared technology is proposed. Infrared visual image detection technology is used to extract the visual characteristics of high voltage insulators during operation. In the image of the high voltage insulator, the pattern, image texture and equipment abnormal state information in the image are analyzed and fused, and the defect detection method is designed. By combining the characteristic quantity of geographic information with the detection system, the defect location of high voltage insulators is realized. At the same time, through the infrared image visual feature point and abnormal feature point calibration technology, the fault, defect and other location information of the high-voltage insulator are calibrated. On the basis of image import and management, defect identification audit, manual labeling and other modules, the intelligent detection of high voltage insulator defect based on infrared technology visual feature recognition is realized. The test results show that the designed intelligent detection method for high voltage insulator defects can accurately calibrate the characteristic quantity of abnormal state information of high voltage insulator, and the root-mean-square error of the detection of defect parts is the lowest 0.061. The reliability and accuracy of the detection of high voltage insulator defects are good.

The machining path has an important effect on the surface quality of complex optical surfaces and the optical properties of the workpiece. Therefore, the optimal path solution method of artificial potential field method is studied to improve the machining quality and meet the practical application requirements. According to the condition of equal residual height, the cutter point of complex optical surface machining is determined. The cutter point is taken as the target point when the artificial potential field method is used to solve the optimal machining path. By introducing the relative velocity repulsion field into the repulsion field function of artificial potential field, the repulsion field function is improved, and the resultant force of each cutter point is determined by using the gravitational force field function and the improved repulsion field function. According to the resultant force of each cutter point, the feed direction of the tool is determined, that is, the optimal machining path of the complex optical surface. Experimental results show that the machining accuracy of this method is 98.15%, which can effectively determine the cutter point of complex optical surface and solve the optimal machining path. The surface roughness of complex optical surfaces machined by this method is lower and the machining quality is higher.

With the gradual improvement of optical instrument accuracy, the frequency of surface dynamic deformation events also shows a sharp increase trend, which is one of the main factors affecting and restricting the development and application of optical instruments. In order to meet the application requirements of optical instruments, a high-precision surface dynamic deformation measurement method for optical instruments based on machine vision technology is proposed. Using machine vision equipment to obtain high-precision optical instrument surface images, the surface images are grayed out, denoised, and enhanced. Based on this, the SURF feature detection algorithm is applied to extract the feature points of the surface image. The FLANN algorithm is used to match the feature points of adjacent images, and Delaunay triangulation is used to form a triangular mesh in the surface image area to further calculate the displacement and strain of the feature points of the surface image, Thus achieving high-precision measurement of surface dynamic deformation of optical instruments. Experimental data show that: When the horizontal coordinate is 1m, the vertical coordinate of the surface dynamic deformation displacement obtained by the proposed method is 1.8m, which is basically consistent with the actual displacement. When the experimental group is 6, the surface dynamic deformation strain value obtained by the proposed method is 3m, which is consistent with the actual strain value, which fully proves that the proposed method has good application performance.

In order to solve the problem of poor visibility of the Vehicle Around View Monitor system affected by ambient light, this paper designs a joint color correction algorithm for multiple original images and a transparency fusion algorithm for adjacent images. Firstly, in the color correction of 4-way fish-eye images, a multi-image joint color correction algorithm is proposed based on the gray world model, which can quickly reduce the color difference caused by illumination of different cameras; Then the affine transformation is realized by using the external matrix of the camera to generate the aerial view; Finally, the illumination adjustment mapping table is introduced into the overlapping area of the aerial view to adjust the edge brightness, and the panoramic 2D aerial view is obtained through image fusion. This paper collects the actual driving video taken by four fish-eye cameras under various light scenes for testing. After processing, the evaluation index of the NQIE of the image in this paper is reduced by 64%, and the running time of a single frame is only 27ms. The results show that the method can stably output panoramic 2D aerial view with uniform illumination under different lighting scenes.