7075 aluminum alloy is a multifunctional industrial metal material with excellent corrosion resistance, mechanical properties and processing properties. It is widely used in aerospace, military and other fields. Focusing on the application of 7075 aluminum alloy additive manufacturing, this paper analyzes the process advantages of laser hybrid additive manufacturing of 7075 aluminum alloy with various heat sources, and discusses the research progress of hybrid technology of 7075 aluminum alloy laser and TIG, MIG and other additive methods at home and abroad. Finally, the problems and future development directions of 7075 aluminum alloy laser hybrid additive manufacturing technology in practical application are summarized.

As a new type of directed energy weapon, laser weapon has gradually stepped into actual combat, and is an important factor affecting modern war in the future. Based on the research results of the US army in the field of laser weapons, the research status, development history, equipment characteristics and key technologies of each platform laser weapons are investigated and sorted out. The future development direction of laser weapon is analyzed, and the data reference is provided for its operational application in the future battlefield.

Conventional microscopy imaging techniques can cause irreversible damage to the activity of transparent cells, while existing light intensity transfer equations lack practical instruments that are both portable and low-cost. To address this challenge, this paper designs a portable microscopic imaging system based on the light intensity transfer equation. According to the requirements of the light intensity transmission equation, the mechanical structure was independently designed and corresponding interactive software was developed. On the premise of ensuring system resolution, it has achieved features such as volume reduction of 70%, high integration, rapid disassembly and assembly, and significant cost reduction. The experimental results demonstrate a resolution of 2.19 microns for the system when using a 10× objective lens. Real time phase quantitative observation of transparent cells can be performed and applied to defect detection of transparent/semi transparent samples. This study provides feasibility for achieving low-cost, portable, and modular design of microscopy imaging technology.

Vortex beam is one of the current research hotspots, with the deepening of research, its transmission characteristics in the medium has also been paid attention to. Based on the nonlinear wave equation, the split-step Fourier algorithm is used to numerically simulate the amplification characteristics of square vortex beams in Nd:YAG media. The effects of topological charge, incident peak light intensity, beam waist width and beam order on the intensity, spectrum and B-integral of a square vortex beam are studied. The numerical results show that when the square vortex beam is amplified by Nd:YAG medium, the intensity fluctuation of the center position is stronger, but the intensity fluctuation of the edge of the beam is weaker. The intensity of the central region of the outgoing beam oscillates continuously, and the intensity in the diagonal direction is stronger than that in the axial direction. Small topological charge and large waist width of the square vortex beam can slow down the central intensity oscillation of the beam through Nd:YAG medium, and reduce the influence of Kerr effect on beam amplification. In addition, reducing the thickness of the Nd:YAG medium can reduce the intensity of the beam center and reduce the size of the B-integral.

Due to its complex spatial distribution, the higher-order Hermite-Gaussian modes has a wide range of applications in precision measurement, optical communications, and laser processing. Although the higher-order Hermite-Gaussian modes can be efficiently generated using cascaded spatial light modulators, the experimental system has extremely high alignment requirements for the optical field and the hologram, which becomes a key factor limiting their high-quality generation. This paper provides a deep analysis of the impact of misalignment between the optical field on the plane of the second spatial light modulator and the hologram loaded onto it. The results indicate that microlevel alignment in the lateral direction and millimeterlevel alignment in the longitudinal direction are required between the optical field on the plane of the second spatial light modulator and the hologram loaded onto it. Additionally, this paper proposes an experimentally simple method to achieve precise alignment between the optical field on the plane of the second spatial light modulator and the hologram loaded onto it, providing important reference for efficiently generating the higher-order Hermite-Gaussian modes with high quality. This research is of great significance for advancing the development of optical imaging, precision measurement, quantum information processing, and other fields.

The utilization of silicon-based slot waveguide structures has garnered significant attention in recent years, particularly in the context of multimode interference couplers (MMIs). In this paper, we present the design of 1×2 parabolic multimode interference power splitter. The coupling zone is set to a parabolic structure, a tapered waveguide is used at the input/output of the coupler. An unequal-width design is used for tapered waveguide at the output, to reduce excess loss. The device is simulated using eigenmode expansion (EME) method. The presented results signify that the device can be fully applied in the C-band and the excess loss is only 0.036 dB, it has better characteristics and can be used in optical network systems.

Using the peak power of 600 W thulium-doped fiber laser, the laser pulse parameters were controlled by Labview program to carry out the extracorporeal lithotripsy experiment, and the law of pulsed thulium laser lithotripsy was studied to provide reference for clinical application. The results show that the pulsed thulium laser has good lithotripsy ability. When the pulsed thulium laser with an energy of 0.3 J is set to impact calcium oxalate stones, most of the debris particles are less than 0.5 mm in diameter, which has good pulverization ability. When the pulse energy is small, the gravel efficiency increases rapidly with the increase of pulse energy. For the thulium laser with a peak power of 600 W, the double pulse combination does not produce obvious ‘Moses effect’ to improve the gravel efficiency. In addition, a large pulse width may be set to improve the efficiency of thulium-doped fiber laser lithotripsy. Under this condition, the penetration distance of laser in water is large, reaching 5.4mm in the experiment. Therefore, the safe distance between the fiber end and the surrounding tissue should be fully considered in the operation to avoid damage to the surrounding tissue.

In order to solve the problems of loss of joint point detection and inability to identify small targets in the scenario of multi-person human pose estimation, an improved YOLOv8-Pose model was proposed. The core improvement of the algorithm is that the convolution in the C2F module is replaced by the variable convolution DCNV2, which enhances the feature extraction ability of the network. At the same time, the weighted bidirectional pyramid BiFPN module is used to replace the feature fusion module in the original model, which aims to retain the small target information and fuse more shallow information to improve the recognition accuracy. Finally, in order to further strengthen the ability to capture and analyze key parts, the SimAM attention mechanism was introduced to weight the local features. Experimental results show that the detection accuracy of the algorithm reaches 70.5% on the CrowdPose dataset, which is 3.3% higher than that of the original model. Compared with the original YOLOv8-Pose model, the improved model not only has higher detection accuracy, but also has a significant improvement in the recognition effect of small targets. It can be seen that the improved network can be applied to multi-person human posture detection more accurately and effectively.

For purpose of improving the accuracy and speed of extracting fringe center of linear structured light, an algorithm based on two-terminal prediction and center intersection interpolation is proposed. Firstly, the extraction region is determined by contour tracking and connected domain analysis algorithm to speed up the calculation and reduce the interference of background information. The center was extracted by gray gravity center method, and the curves were fitted to the extracted results. Then, from the center point of the linear structured light fringe to both ends, the initial extraction is carried out: the distance from the extraction point to the fitted curve is calculated, and the average distance is taken as the threshold, and the two-terminal prediction is used when the distance is greater than the threshold. Then, the normal direction of each position of line structure light fringe is calculated by block principal component analysis method, and the interpolation point is obtained by central intersection interpolation method. Finally, the gray gravity center method is applied to the interpolation points to obtain the accurate fringe center coordinates of line structure light. The experimental results show that the proposed algorithm can accurately extract the fringe centers of linear structured light at places with large curvature changes and strong reflections. When extracting the fringe centers of linear structured light generated by computer simulation, the extraction accuracy of the proposed algorithm reaches 64.1 times that of the gray gravity center method and 6.0 times that of the Steger Algorithm, and the running speed of the proposed algorithm is slightly slower than that of the gray gravity center method and 7.6 times that of Steger algorithm.

The paper proposes a DRM-YOLOv8n small target detection algorithm to enhance the accuracy of image detection for PCB defect detection. Firstly, the deformable convolutional module is employed to address the issue of small-scale feature extraction, thereby enhancing the backbone network’s capability in extracting crucial features and improving detection accuracy. Secondly, a receptive field attention Convolution (RFAConv) is introduced into the Neck structure to improve model positioning accuracy in complex scenes and enhance network performance and efficiency. Lastly, the MPDIoU loss function is utilized to optimize the original network loss function, resulting in improved boundary box regression accuracy and model convergence ability. Compared with YOLOv8n, our proposed algorithm achieves significant improvements in average precision value (mAP), with mAP50% increasing from 87.2% to 94.5% and mAP50:95% increasing from 60.9% to 65.8%, respectively - representing a 7.3% and 4.9% improvement over YOLOV8n.

A multi cable fault automatic synchronous detection method based on bidirectional LSTM and laser scattering signals is proposed to address the current issue of poor detection performance for multiple cable faults. Firstly, collect laser scattering signals from multiple cables simultaneously through laser scattering equipment; Secondly, the waveform characteristic data of the cable laser scattering signal is obtained through nonlinear matching tracking method, and this is used as input for the cable fault detection model; Finally, attention mechanism is introduced into the bidirectional LSTM model to achieve automated synchronous detection of multiple cable faults based on the improved bidirectional LSTM model. The experimental results show that the proposed method has higher accuracy and precision in automatic synchronous detection of multiple cable faults, and is more suitable for practical applications.

In response to the demand for precise birefringence measurements, this study investigates the utilization of double photoelastic modulation for birefringence measurements. The primary focus lies in addressing challenges related to online, in-situ calibration of the photoelastic modulator (PEM) and data processing within the measurement setup. Digital phase-locking technology is employed to facilitate the assessment of multiple modulation signal frequency components. The amplitude ratio of each frequency modulation signal is employed for online, in-situ calibration of PEM phase amplitude. Furthermore, leveraging this amplitude ratio enables the determination of birefringent retardance and fast-axis azimuthal angle measurements. An experimental setup was constructed and experimentation conducted using Solier’s Barbierne compensator as a standard sample. Results demonstrate that the proposed approach enables real-time online, in-situ calibration of PEM phase amplitude, facilitating high-precision, high-sensitivity, and rapid birefringence measurements. The relative error remains below 0.60%, and standard deviation doesn’t exceed 0.049 nm within a 200 ms timeframe. This method exhibits promising prospects for the real-time detection of birefringence in optical materials and components, offering significant implications for practical applications in online birefringence detection.

CO and CO2, as important products in the combustion process, can increase the pressure of carbon reduction when they are emitted in large quantities. Therefore, high-precision real-time online monitoring of CO and CO2 concentrations is of great significance for improving boiler combustion efficiency and accurate measurement of carbon emissions. In this work, an online measurement experimental system based on wavelength modulation direct absorption spectroscopy (WM-DAS) technology and Herriott multiple reflection cell was designed. This experimental system can achieve accurate concentrations of CO and CO2 in mixed gases of the order of 10-6. And the system detection limit can reach the order of 10-8. Then, based on the experimental system, experiments were conducted to study the high-temperature catalytic oxidation characteristics of CO. The effects of oxygen content and CO concentration on catalytic efficiency were investigated. The experimental results showed that under the conditions of temperature ranging from 339 to 344 K and pressure of 101 kPa, the CO catalytic efficiency increased with the increase of oxygen content in the range of 1% to 5%. When the oxygen content was 1% and the CO volume fraction was 800×10-6, the catalytic efficiency increased with the increase of reaction temperature. The experimental results can provide a data basis for the removal and conversion of CO in industrial flue gas.

The Laser-induced breakdown spectroscopy and spatial confinement method were used to study the optimal detection delay time of the characteristic spectral lines of different elements, to establish a plasma thermal expansion model, to simulate the thermal expansion transient process of the plasma at different spatial confinement depths and times, and to analyze the effects of changes in ambient air pressure, temperature and humidity on the results. Laser-induced breakdown spectroscopy test experiments were carried out on the samples of the traditional Chinese medicine yucca with different spatial confinement depths, using Mg Ⅰ 285.12 nm and Ca Ⅱ 396.84 nm as the characterized spectral lines to analyze the effects of changes in spatial confinement depth and time on the intensity of the spectral line, as well as the effects of changes in excitation light output characteristics on the spectral intensities were also explored. The spectral line intensities were significantly enhanced at a laser energy of 30 mJ, a repetition frequency of 1 Hz, and an optimal spatial confinement depth of 1.9 mm, and the optimal detection times were 0.5 s for Mg and 1.2 s for Ca, which indicated that the optimal detection times for different elements with the same spatial confinement depths were different for the enhancement of the spectral lines.

To address the catastrophic forgetting problem in lifelong learning, this paper proposes a lifelong learning method for hyperspectral image classification using multi-level knowledge distillation. Initially, a feature extractor based on multi-modal alignment is designed to fully leverage the spatial-spectral information and label text information of hyperspectral images. Additionally, a multi-level knowledge distillation strategy is devised to effectively preserve the multi-modal knowledge from previous phases. The proposed method was experimented on two public hyperspectral datasets. Compared to current state-of-the-art methods, the proposed method showed an average accuracy improvement of 15%-18% on the Pavia University dataset and 1%-8% on the Botswana dataset.

When unmanned detection platforms utilize the complementary advantages of LiDAR and visible light cameras, the effect of calibrating and fusing point clouds and color images solely based on feature points on the calibration object is not ideal, and multiple data collections are required to fit the results. A fusion method based on distributed automatic calibration of LiDAR and visible light camera is proposed to address the above issues, which enables the fused data to possess both the spatial stereoscopic characteristics of point clouds and the color texture information of visible light images. This method first calibrates the laser radar and camera through a planar calibration board to obtain the initial transformation relationship, then aligns them using the natural edge features in the common field of view of the two sensors, and finally fuses the two data to obtain a visualization result. Compared with the calibration fusion method based on manual matching and the automatic calibration fusion method based on planar calibration plates, the proposed method has improved accuracy by 33.8% and 23.1%, respectively, and the visualization results effectively restore the real scene.

Large scale laser image recognition methods are limited by computational power and data processing speed, making it difficult to meet the needs of fast classification and recognition. The emergence of cloud computing provides ideas for solving this problem. Design a fast classification and recognition method for large-scale laser images under cloud computing. Using cloud computing platforms to perform mathematical morphology analysis and preprocessing on laser images, ensuring that the images maintain consistent morphology and features in subsequent processing. With the help of wavelet analysis technology on the platform, the preprocessed laser images are subjected to denoising processing, effectively removing noise interference in the images and improving image quality. After obtaining noise free images, the distributed resources of cloud computing platforms are further utilized to efficiently extract image features and obtain feature parameters that accurately describe image texture attributes. Combine these feature parameters with other features to construct a complete image feature vector. By comparing the feature vectors of different images, accurately determining their similarities and differences, achieving precise classification and recognition of laser images. Comparative analysis of experimental data shows that compared to traditional methods, this method has significantly improved processing speed and recognition accuracy. It can not only effectively process large-scale laser image data, but also significantly improve the accuracy and efficiency of image recognition, providing new ideas and methods for the development of laser image recognition technology.

Accurately capturing the geometric shape and structural information of buildings is crucial for point cloud reconstruction of damaged building scenes. Aiming at the problem of low completeness in point cloud reconstruction of damaged building scenes, a point cloud reconstruction method for damaged building scenes is proposed by integrating line segment descriptors and beam adjustment. Using the Mask RCNN method to detect damaged building scene targets and implementing two-dimensional projection on the obtained point cloud plane. Using principal component analysis method to estimate all point cloud normal vectors, introducing region growth method to cluster and segment all point cloud data, fitting each plane in the damaged building scene image, and obtaining the point cloud plane. Using the two-dimensional line segment detection method to obtain the line segment features of the damaged building plane, projecting the two-dimensional line segment into the three-dimensional space to obtain the three-dimensional line segment, and generating the line segment descriptor corresponding to the damaged building scene. Implement beam adjustment optimization on the obtained line segment descriptors to achieve point cloud reconstruction of damaged building scenes. The experimental results show that the proposed method can effectively reconstruct the point cloud of damaged building scenes, and the minimum relative error of point cloud reconstruction is only 0.117.

Image is one of the key information carriers in the field of visual communication at present. Due to the influence of many factors such as acquisition equipment, compression technology, transmission environment and display mode, the clarity of visual communication image is poor, which restricts the communication effect of visual information. Therefore, a clear processing method of visual communication image based on laser visual data fusion is proposed. The laser radar and monocular camera are used to obtain the relevant information of the visual communication object, and the laser data features, such as point cloud density, laser data point normals and laser data point curvatures, and visual data features, such as color features, texture features, shape features and spatial relationship features, are extracted. On this basis, a laser vision data fusion framework is established, the fusion scale of laser data features and visual data features is determined, and the laser data is projected onto the visual image, thus realizing the clear processing of the visual communication image. The experimental results show that the visual communication image obtained by the proposed method has high definition, and the maximum information entropy of the visual communication image is 35 bit, which fully proves that the proposed method has better image clarity effect.

To address the loss of details, low brightness, and contrast in existing fusion algorithms, the article proposes the Fusion of Low-Light visible light and infrared images based on Parallel Networks (PNLLFusion). PNLLFusion aims to maximize the preservation of detail information from the source images and enhance brightness and contrast. This method implements parallel fusion and brightness enhancement, reducing information loss caused by incompatibility between enhancement and fusion algorithms. Additionally, residual structures are added on the Squeeze-and-Excitation (SE) module, and gradient computation is incorporated into the self-attention network to preserve more texture and edge information. The effectiveness of the method is validated on the LLVIP dataset and TNO dataset. Experimental results demonstrate that compared to classical fusion algorithms, this method can preserve more detail information from the source images in low-light environments, while also improving image contrast and brightness. It achieves good or comparable results in both subjective and objective evaluations.

Aiming at the difficulty of target recognition in infrared images, a method for identifying weak and small moving targets in infrared images based on background difference method is proposed. Firstly, a spatiotemporal regularization correlation filtering algorithm is used to process a large range of background areas near the target, and side window filtering is used to further remove noise, denoise and preserve edges in low-quality infrared images, and improve image quality. Then, the background difference method is used to preliminarily identify weak and small targets in the image, Finally, based on the preliminary recognition results, Kalman filtering is introduced to predict the motion trajectory position of weak targets in infrared images, achieving weak motion target recognition. The experimental results show that the proposed method has a higher recognition success rate and more accurate recognition position.

A laser digital image block encryption method based on an improved Clifford chaotic system is proposed to address the serious problem of sensitive information leakage caused by frequent attacks on laser digital images during transmission and storage. Firstly, the Clifford chaotic system optimized by parameters is used to block and scramble laser digital images. By shuffling the arrangement order of image pixels, the complexity of image information and the ability to resist attacks are effectively increased. Secondly, the Logistic chaotic mapping algorithm is used to encrypt the scrambled image sub blocks, ensuring that the information within each sub block is sufficiently confused and diffused. Finally, the overall security of the image is further improved by recombining and re encrypting the encrypted sub blocks using double random phase encoding technology. The experimental results show that this method performs well in protecting laser digital image information, and the encrypted image information has high security, which can effectively resist various attack methods. Meanwhile, the encryption effect of this method is good, and the encryption process is efficient and stable.

Due to the influence of external environment, LiDAR images are easily affected by various noises, which reduces the accuracy of data. To this end, a sparse denoising method for LiDAR images based on deep learning is proposed. We use an accelerated backward projection algorithm to generate initial LiDAR images. In response to the image blurring phenomenon generated during the imaging process, we set adaptive transition points and enhance blurry contrast to complete the deblurring processing of the LiDAR images. Combining the advantages of deep learning technology, an adaptive stack style sparse denoising autoencoder is established. Through multi-channel SRDA, each SDA is trained for different types of noise, and finally linearly combined to handle multiple types of noise simultaneously. This multi-channel approach can more comprehensively eliminate various noises and improve the sparse noise reduction effect of LiDAR images. The experimental results show that the proposed method not only effectively removes the blurring phenomenon of LiDAR images, but also has a relatively efficient denoising ability.

To address the challenge of accurately estimating the 6D pose of satellites in space environments characterized by significant lighting variations and complex background changes, as well as limited satellite texture features, a novel approach is proposed. This method combines partial convolution (PConv) and large kernel attention (LKA) within the framework of the DenseFusion network. Firstly, improvements are made to the generation of Blender-based rendering datasets and a dedicated simulation dataset for satellite pose estimation is created. Subsequently, the integration of a partial convolution module into the feature extraction network’s encoding section reduces sensitivity to lighting changes and background noise. Finally, to capture weak texture features at different scales on satellite images, a pyramid scene parsing network LKA-PSPNet (Large Kernel Attention Pyramid Scene Parsing Network) is designed. Experimental results demonstrate that this algorithm achieves an ADD-(S) index of 97.6% and 89.2% on both LineMod public dataset and self-built satellite simulation dataset respectively- marking an improvement by 3.3 percentage points and 2.9 percentage points over previous methods- thus validating its effectiveness.

The data quality of fiber optic networks can affect the communication quality of users. When abnormal data is included, it will significantly reduce communication efficiency and quality. To this end, a multi-channel fiber optic network big data anomaly detection algorithm based on high-dimensional spatial clustering is proposed. Obtain abnormal data features based on historical data feature density indicators; Using high-dimensional Gaussian mixture clustering algorithm to map data features from low dimensional space to high-dimensional space, in order to reduce computational difficulty, using kernel mapping to convert high-dimensional inner product calculation into low dimensional data kernel calculation; Finally, the HGMM algorithm is used to obtain the detection time series at different times, and the final detection result is output after fusion. The experimental results show that the proposed method can achieve a feedback rate of up to 98%, and the detected abnormal data types are completely consistent with the actual results, ensuring that the fiber optic network is not affected by abnormal data.

As one of the current key communication technologies, LiDAR network electronic communication faces an increasing level of security risks as its application scope continues to expand, resulting in the inability to effectively guarantee the security of electronic communication data transmission. Therefore, this article proposes a secure transmission method for electronic communication data in LiDAR networks based on symmetric encryption. This method introduces symmetric encryption algorithms to build a secure transmission framework for electronic communication data in LiDAR networks. Under this framework, electronic communication data is encrypted through data processing, byte substitution, row shifting, column obfuscation, and round key addition, while determining the key extension mode. The decryption process is completed through ciphertext processing, reverse row shifting, reverse byte substitution, reverse column obfuscation, and reverse round key addition, thereby achieving secure transmission of electronic communication data. The experimental results show that the minimum secure transmission time of electronic communication data obtained by the proposed method is 1 second, and the minimum transmission loss rate of electronic communication data is 0.6%, fully confirming that the proposed method has better application performance.

With the rapid development of modern optical communication networks, the security of optical networks has attracted much attention. In order to ensure the accuracy of information and reduce the frequency of abnormal information in optical networks, a particle swarm optimization based method for extracting abnormal signal data from optical network nodes is proposed. Firstly, design a filter bank to perform frequency band processing on the signals of optical network nodes. On this basis, construct decision statistics and use them as the basis for decision processing of each frequency band to achieve node signal enhancement processing; Secondly, based on the empirical modal algorithm, the node signal is decomposed, and effective IMF components are obtained through screening, and their energy is calculated as the signal feature of the node, providing a basis for the subsequent extraction of abnormal data. Finally, the particle swarm optimization algorithm is used to optimize the weights of the BP neural network, and the node signal features are input into the optimized neural network to extract abnormal data from the optical network node signals. Experimental verification shows that this method has a good effect on enhancing node signals, with high accuracy and stability in extracting IMF component energy and abnormal data of optical network node signals.

Design optical communication coding schemes under time-varying channels to promote optical communication technology in reducing inter symbol interference and channel distortion in complex and changing channel environments, achieving more efficient and stable data transmission. Construct a time-varying optical communication channel model that considers factors such as multipath effects, nonlinear distortion, signal attenuation, and noise, and introduce direct sequence spread spectrum (DSSS) technology and parameterized inverse time matrix (PTRM) technology to mitigate inter symbol interference and channel distortion; At the transmitting end of optical communication, an optical communication encoding scheme was designed using low-density parity check (LDPC) encoding technology. By constructing a sparse parity check matrix and using the row column permutation algorithm (RU algorithm) to convert the parity check matrix into an approximate lower triangular form, an LDPC encoded codeword vector was generated; At the receiving end, a soft decision decoding method is used to gradually approximate and determine the transmission codeword through fine multi-level quantization processing and iterative algorithms. The experimental results show that under different temperature fluctuations and signal-to-noise ratio conditions, this encoding scheme can significantly reduce the bit error rate.

In order to achieve precise intrusion detection in fiber optic networks, a fiber optic network intrusion detection method based on Markov decision process is proposed. By using frequency domain partitioning technology to purify fiber optic network signals, the empirical mode decomposition method is used for initial detection of intrusion signals, and the fuzzy analytic hierarchy process is used to determine the credibility of network access behavior. For access behaviors with higher credibility, they are directly passed through, while the remaining access behaviors are judged using Markov decision process, thus achieving intrusion detection. The experimental results show that this method can quickly and accurately detect intrusion signals, especially for the intrusion eavesdropping behavior suffered by the lending dataset, with a detection rate of up to 0.985. In the entire experiment, the minimum detection rate of this method can also reach 0.920, and the maximum average detection misjudgment rate and average detection omission rate are 0.01 and 0.02, respectively. This indicates that the method significantly improves the security and stability of fiber optic networks, providing strong support for ensuring network security.

Laser induced breakdown spectroscopy (LIBS) was used to ablate aluminum samples, and the influence of 1 064 nm and 532 nm lasers on plasma emission spectra were studied. The influence of different laser wavelengths and energies on spectral signal intensity were compared, it was found that 1 064 nm was more likely to excite non-metallic element spectral lines, while 532 nm was more likely to excite metallic element atomic spectral lines. The influence of different energies and wavelengths on signal stability were studied, it was found that short wavelength (532 nm) signals have better stability, with an average relative standard deviation (RSD) of 0.72 times that of 1 064 nm lasers. At the same time, the relationship between laser wavelength and spectral energy level difference has been studied, and it was found that the relative intensity of spectral signals generated at 532 nm and 1 064 nm is negatively correlated with their energy level difference. Appropriate selection of laser wavelength and energy can effectively improve signal sensitivity and stability.

Traditional interactive virtual display technology suffers from poor interactive performance in special viewpoints of dynamic scenes. Therefore, an interactive virtual display technology based on light space transformation technology is proposed. An interactive virtual space is established based on the light space transformation technology. The process of converting 2D to 3D stereoscopic scenes is analyzed, and the key points controlling the overall shape and the left view are extracted. The coordinate information of the light space interaction points is derived. Gaussian blur processing is applied to the images in the modeling process. The processed images are secondarily matched with the interaction point regions to optimize the design process of the interactive virtual display. The simulation experiment results indicate that among the selected four groups of test data, compared with several traditional interaction methods, the designed method features the best recognition effect, the shortest time consumption, the highest efficiency, and the most stable comprehensive performance, demonstrating favorable application effects.

To improve the sensitivity, resolution, and other indicators of laser sensors, a mathematical model for optimizing laser sensor parameters based on data mining is constructed., Obtain dynamic parameters such as the moving distance and different angles of the target object within the measurement range of the laser sensor, and then use the EMD algorithm to mine the feature vectors of the dynamic data during the measurement of the laser sensor. The RBF neural network judges whether the current laser sensor parameters deviate from the theoretical values in the normal state based on the feature vectors, establishes a mathematical model for optimizing the laser sensor parameters, and optimizes the laser sensor parameters that deviate from the normal state. The experiment shows that this method can accurately distinguish whether there is a deviation in the distance between the laser sensor emitter and the object image reference, and can effectively optimize the parameters of the laser sensor. The sensitivity has been improved by 15.29%, and the application effect is good.

In recent years, frequent fire has caused a large number of casualties and property losses, so fire warning is becoming more and more critical. In order to effectively solve the fire early warning system response is relatively slow and the detection range is limited, resulting in limited early warning ability. Combined with the characteristics of fire, Jetson Nano (artificial intelligence edge computing suite) was used as the main control center, and LabelImg (label) was used to label more than two thousand image data. The training weight of fire image was obtained by YOLOv5 (target detection algorithm) training. Jetson Nano reads the image data collected by the UV camera and calculates whether there is a fire by training the weight. If there is a fire, it sends data to the microcontroller through the serial port of Jetson Nano to trigger the alarm. This design has been tested for fire and smoke detection accuracy of more than 80%, smoke and fire identification accuracy is high, fire early warning is rapid. Compared with the traditional fire detection technology, the detection range is wider, the accuracy is higher, and it is not easy to be affected by the environment.

Microcracks are defects that may occur in stainless steel sheets during use, and these cracks may affect the strength and stability of the material. The current detection methods are affected by interference sources such as sound, vibration, or electromagnetic waves in the surrounding environment when detecting these microcracks. The generated signals will be superimposed on the diffracted wave signal, causing the interference signal to mix with the diffracted wave signal, making it difficult to accurately extract the characteristics of the diffracted wave signal, thereby affecting the accuracy of crack angle detection. To this end, a laser ultrasonic diffraction wave detection technology for internal micro crack angles in stainless steel plates is proposed. Using laser ultrasound technology to detect the interior of stainless steel plates, generating ultrasound waves through laser excitation, and obtaining diffraction wave signals caused by microcracks. Perform anti-interference processing on diffracted wave signals to eliminate mixed interference noise in the diffracted wave signals. Conduct in-depth analysis on the diffraction wave signal processed by anti-interference, identify and extract features related to micro crack angle detection - diffraction wave propagation direction, diffraction wave energy distribution, amplitude, phase, and propagation time. Based on these features, a calculation formula for microcrack angle is constructed to achieve the calculation and detection of microcrack angle. The experimental results show that the maximum error between the microcrack Angle detection results obtained by this technique and the actual results is only 1.1°, and the shortest detection time is only 11.8 min, which fully confirms the superiority and high efficiency of the proposed technique.

Since the vehicle laser point cloud data usually contain a large number of points, including noise points, and directly calculate the adjacent points of each point, the road surface point cloud extraction is not accurate, resulting in low detection accuracy. Therefore, a pothole detection method based on vehicle-mounted laser point cloud data was studied. Firstly, the KD tree algorithm and the CFS algorithm optimized by least square method are used to filter the laser point cloud data of the vehicle, avoid directly calculating the adjacent points of each point, reduce noise and improve fitting accuracy, so as to accurately extract the road surface point cloud, and map the road surface point cloud to the XOY plane through mapping processing to generate road surface images. The low frequency part of the image is processed by gamma correction, the high frequency part of the image is processed by bilateral filtering, and the enhanced image is obtained by fusion of the above processing results. The pothole area is determined by the characteristics of the connected domain, and the pothole depth is calculated to realize the pothole detection. The experimental results show that the proposed method can effectively extract the pavement point cloud, the pavement image processing effect is good, and 5 pavement potholes are accurately detected, and the shape of the detected potholes is basically consistent with the actual shape, and the position deviation is small, which has a good image processing effect and accurate detection result.

In the layout process of LiDAR and camera, the original layout constraints are not precise or comprehensive enough, resulting in lower coverage after layout, which affects the reliability of measurement results. Therefore, a laser radar and camera layout optimization method based on hot spot areas is proposed. Complete the motion trajectory data of the target area, preprocess it, obtain scene hotspot data, and determine the scene hotspot area. Based on the obtained scene hotspot data, laser radar layout constraints are constructed through three parts: distance, angle, and collision. Determine the motion speed of the LiDAR, use a grid to partition the hot spot areas of the scene, and apply constraint conditions to solve for the optimal layout position. Constructing an experimental section, the experimental results show that this method achieves layout optimization and can further improve the coverage of LiDAR and camera layout, ensuring the reliability of LiDAR and camera measurement results.

The error of laser sensors directly affects their accuracy performance. In order to provide data support for the calibration of laser sensors, big data analysis technology is used to optimize the design of real-time estimation methods for laser sensor measurement errors. Taking into account the composition, structure, and working principle of laser sensors, simulate the measurement process of laser sensors. Determine the influencing factors and mechanisms of measurement errors from both internal and external perspectives, and use the collection and cleaning techniques in big data analysis to collect the working data and working environment data of laser sensors. Obtain the estimated results of measurement errors of laser sensors at any time. By comparing with traditional methods, it can be concluded that the estimated deviation value of the measurement error of the laser sensor optimized by the design method is reduced by about 16.5 mm.