Laser Induced Fluorescence (LIF) technology has the advantages of high sensitivity, high resolution, and non-destructive properties, and is currently a hot research topic. The continuous development and application of LIF technology provide us with more possibilities to deeply understand and solve important issues such as environmental pollution, food safety, and disease diagnosis, and make positive contributions to achieving sustainable development and human health. This article mainly introduces the principle, detection system, and classification of LIF technology. It reviews the current application of laser induced fluorescence spectroscopy in the detection of polycyclic aromatic hydrocarbons (PAHs) in soil, food safety detection, and biomedical detection. Finally, it looks forward to the development trend of LIF technology in the field of detection, laying a foundation for further development of LIF technology in soil pollution, food safety, and biomedical detection in the future.

To address the issue of diminished quality in bidirectional controlled quantum teleportation between space and ground during rainy conditions, this paper proposes a scheme based on a ten-particle entangled state. In this scheme, communication entities Alice and Bob are capable of simultaneously transmitting arbitrary two-particle quantum states bidirectionally, contingent upon approval from the satellite’s control unit, Charlie. Furthermore, to investigate the impact of rainy conditions on the performance of this teleportation scheme, taking amplitude damping noise channels as an example, the study analyzes the fidelity and influencing factors of bidirectional controlled quantum entanglement-based teleportation between space and ground. Lastly, to counteract the effects of rainy conditions, weak measurement and reversal measurement operations are employed to safeguard quantum entanglement. Theoretical analysis and simulation results demonstrate that weak measurement and reversal measurement operations can effectively enhance the fidelity of space-to-ground teleportation, highlighting the potential value of the approach presented in this paper for improving the quality of controlled quantum teleportation in rainy environments. This research offers valuable insights and methodologies for the application of quantum communication in challenging environmental conditions.

A balanced multistage low-noise amplifier is proposed to address the problem that the acquisition of backscattered light mode signals in the backscattered light measurement of laser gyro resonance cavities requires low-noise amplifiers in specific frequency bands. The low-noise amplifier adopts a balanced multistage cascade structure, designs a special orthogonal coupler to achieve conjugate matching, adjusts stability based on negative feedback, optimizes noise for the bias circuit, selects impedance points based on noise and gain, and analyzes and innovates the structure of the orthogonal coupler from the perspective of power flow transmission. The design goal of " impedance matching and noise minimization" is finally achieved. Verification shows that the gain is up to 20dB at 2.3 GHz~2.7 GHz, the gain flatness is about 3dB, the overall input VSWR is less than 1.50, and the output noise of the photodetector is obviously suppressed. The design results can be referenced and applied to many scenarios such as signal amplification of high-frequency optoelectronic devices, mixer design, and TR components to solve the problem of low gain and high noise of single-branch low-noise amplifiers.

Experimental research was conducted on the selection of radiofrequency field intensity in Mx-type radiofrequency-driven magnetic resonance magnetometer measurements. The influence of radiofrequency field intensity on the characteristics of resonance signals was investigated using a cesium atomic magnetometer system in an unmagnetized environment. Measurements were carried out using a frequency scanning method, and features such as signal amplitude, linewidth, and slope were analyzed. The results indicate that the optimal radiofrequency field intensity accounts for approximately 3.6% to 4.1% of the magnitude of the magnetic field being measured, at which point the amplitude of the signal resonance peak reaches its maximum value. Additionally, it was observed that the slope of the phase signal resonance peak varies directly with the radiofrequency field intensity, leading to the deduction that the system relaxation rate is inversely proportional to the radiofrequency field intensity. This research provides important guidance for determining suitable radiofrequency field intensities and offers innovative insights into magnetic resonance magnetometer measurements.

Panoramic cameras, with their broad field of view and rich environmental information, are often employed for visual guidance and positioning of extravehicular space robots. To provide high-precision structural parameters for panoramic vision positioning, this paper proposes a calibration method for the structural parameters of a multilens combined panoramic camera system. High-precision photogrammetric measurements of a constructed calibration field are taken, and by utilizing the coordinates of markers in the world coordinate system and their corresponding image plane coordinates, the absolute poses of four cameras are determined. Subsequently, constraints are established based on the relationship between the origin of the world coordinate system and the positions of each camera, to solve for the pose transformation relationships between the reference camera and the other three cameras, which represent the system’s structural parameters. Experimental results show that the standard deviation of the image plane re-projection error of markers for all cameras in the panoramic camera system is 0.827 5 m (1/9 Pixel), with a relative positioning error for all structural parameters of 1.03 mm, and a relative angle calibration error of 0.03 rad. This method can meet the high-precision requirements of structural parameters for the panoramic vision positioning of extravehicular space robots.

Pulse laser ranging is widely used in remote sensing, mapping, control and navigation, detection, recognition and tracking of military targets. The transmitting and receiving system directly determines the ranging capability of the equipment. To meet the requirements of long-distance and eye-safe, the transmitting and receiving system are designed based on the 1.535 m Erbium glass laser. The transmitting optical system used aspherical surface to reduce lenses makes the divergence angle less than 0.4 mard. The receiving system adopts the telephoto mode to reduce the spherical aberration and the length of the system and the coupling efficiency is greater than 90%. The theoretical calculation shows that the systems can realize long-distance ranging, and the structure is very compact, which have high engineering application value.

The alignment accuracy of the two-mirror telescopes is a key link to ensure the imaging quality of the system. One of the currently used alignment methods is to use the coma of the on-axis field of view(FOV)as a reference, which cannot effectively guarantee the imaging quality of the system’s entire FOV. Based on the vector wave aberration theory, the aberration field characteristics of the two-mirror telescopes is analyzed in this paper. According to the characteristics of the coma field and astigmatism field of the misalignment system, an alignment method based on the astigmatism of symmetric off-axis FOV is proposed. According to the theoretical analysis results, taking an R-C two-mirror telescope as an example, the alignment method using on-axis field coma and the off-axis symmetric field astigmatism proposed in this article were analyzed under a given initial misalignment state. A systematic simulation alignment experiment was carried out based on the method in the paper. The results showed that when the on-axis coma is zero after moving the secondary mirror, the astigmatism of the full FOV was 1.02, and the system did not reach the optimal alignment state. However, after the alignment using the astigmatism of off-axis symmetrical FOVs, the average astigmatism of the system’s entire FOV is 0.46, which is basically consistent with the theoretical design value, indicating that the method proposed in this article is effectively completed the alignment of a two-mirror telescope.

Due to the traditional machine vision algorithms there are small feature defect targets on the surface of integrated circuit chips are not obvious to detect, the detection speed is slow and the false detection rate is more. Based on the above problems from both accuracy and efficiency considerations, an improved Faster RCNN namely VanillaNet-11 minimalist network model is proposed as the backbone architecture, avoiding the traditional algorithm ResNet-50 to bring deep and complex link and attention mechanism problems, while further improving the accuracy. The original ROI Pooling layer is replaced by ROI Align to quantify the mismatch problem. Finally, the original NMS is improved by PSRR (Pyramid Shifted with Relationship Recovery) -Maxpool NMS method, which achieves the effect of suppressing the redundant overlapping candidate frames faster while guaranteeing the accuracy. The experimental results show that the average accuracy of the improved network reaches more than 95% in the collected chip defect dataset, which is about 25% higher than that of the original Faster RCNN network, and effectively improves the detection capability of small defective targets.

In order to solve problem of measuring the inner surface roughness and realize the full surface roughness measurement, a white light confocal method for measuring the full surface roughness is proposed. Based on the white light confocal sensor, the measurement system is built, and the workpiece clamping structure, the measured surface positioning structure, the sensor focusing structure and other structure are designed. The software system is developed on the C# platform, and the full surface roughness measurement ability is estabished. Verification experiments have been carried out on the standard multi-scribed template and curved surface samples. The experimental results show that the measuerment system has high indication accuracy, repeatability and stability, and has the ability to measure the full surface roughness of curved surface parts.

With the development of laser technology, the peak power of laser pulse reaches 1012 W, and the pulse duration reaches femtosecond or even attosecond level. The emergence of ultra-short and ultra-intense laser pulse technology has brought great challenges and development opportunities to the military field, especially the mid-infrared band laser pulse has a huge application prospect in laser warning technology, laser detection, laser active guidance, etc. However, due to the insensitivity of the existing detection technology to femtosecond and attosecond laser pulses, a new femtosecond laser pulse detection technology is urgently needed, this paper proposes a super-resolution coherent detection method for mid-infrared band laser pulses and conducts an in-depth analysis of its detection mechanism, which has the advantages of high detection accuracy and high sensitivity, and can be applied to the laser guidance technology with femtosecond or even attosecond pulses as the active light source in the future, which lays a technical foundation for the application of femtosecond and attosecond laser pulses in the military field.

Aiming at the problems of low recognition rate and poor robustness of tea buds in complex backgrounds, an improved YOLOv8 tea bud detection algorithm is proposed. By introducing the Swin Transformer self attention mechanism, a CTS feature extraction module is constructed to enhance the global feature extraction capability of the model; Drawing on the idea of multi-scale fusion, constructing the ExFModule module enriches semantic feature information while enabling the network to adaptively select useful features and suppress useless ones; In terms of feature fusion, a BFPAN feature map stitching method is proposed to enable the model to pay more attention to small target features and improve its feature fusion ability. The experimental results show that the improved YOLOv8 algorithm achieves an average accuracy of 93.4% on the tea tender bud dataset, an increase of 4.4% compared to before improvement, and the detection speed remains basically unchanged. It can achieve fast and accurate tea tender bud recognition and detection, and provide technical support for the intelligent picking of tea tender buds.

Aiming at the problem of missed and false detections caused by the lack of semantic information in shallow feature layers and the lack of detailed information in high-level feature layers in the SSD (Single Shot Multibox Detector) object detection algorithm, an improved SSD object detection algorithm is proposed. First, the improved Comprehensive Convolutional Block Attention Module (CCBAM) is introduced to increase the sensitivity of the network to small targets. Then, a Hierarchical Feature Fusion Network (HFFNet) is constructed to fully integrate the detailed information from shallow layers and the semantic information from high-level layers. In addition, dilated convolutions are used during downsampling to extract features of different scales. In the upsampling process, pixel shuffling is used to increase the resolution of the high-level feature layers while reducing information loss, and then fused with the low -level feature layers to enhance the semantic information in the low-level feature layers. Finally, the Residual Feature Fusion Module (RFFM) is used to improve the integration of local and global information in the high-level feature layers and to enrich the feature information. The experiment demonstrated a 2.4% improvement over the original SSD algorithm, achieving a mAP@ 0.5 of 79.6% on the PASCAL VOC2007 test set at a recognition speed of 47.8 FPS.

Aiming at the problem that there is no applicable flexible wire harness recognition method in the current wiring automation production process, a flexible wire harness recognition method for wiring robots is proposed. Firstly, based on the UNet network structure, ResNet-34 is adopted as the feature extraction network in the encoder, and the SE attention module is introduced in the decoder part to construct a new image segmentation model RS-UNet for the segmentation of wire harness images. Then, the Zhang-Suen image refinement algorithm is used to refine the wire harness mask image, and finally the geometric center position information of the wire harness is obtained to give guidance to the wiring robot to operate the wire harness. Through experiments, it is proved that the RS-UNet network improves the IoU value by 4.95% and the F1 value by 0.029 in the wire harness segmentation accuracy compared to the when -UNet network, and the selected Zhang-Suen refinement algorithm has an average processing time of 0.38 s, an average refinement sensitivity of the image refinement result of 0.39, and an average thickness reference value of 0.87, the proposed method can accurately identify the geometric center of the flexible wire harness.

Aiming at the problems of the current real-time semantic segmentation algorithms based on clothing segmentation (such as Deeplab series, ERFNet series, BiseNet, etc. ), such as insufficient detail features, poor segmentation accuracy, slow reasoning speed, etc., a deep dual-resolution fusion multi-scale real-time semantic segmentation algorithm MDRNet was proposed. In order to improve the learning ability of clothing image features, a dual -resolution feature network was used to extract high-resolution and low-resolution features simultaneously, and a bilateral fusion module was used to integrate features of different resolutions for several times, thus improving the deep supervision output structure of the high-resolution network, and a cross-space multi-scale attention module was added to capture cross-space feature information. Finally, the context extraction module DAPPM was changed to the parallel module PAPPM, which not only improved the accuracy, but also effectively reduced the number of parameters and computational complexity, and effectively improved the FPS. The experimental results showed that MDRNet improved the average crossover ratio of DeeplabV3+and BiseNetv2 algorithms by 9.2% and 8.1% respectively on the DeepFashion-MultiModal dataset for clothing segmentation, providing a better technical solution for the application of semantic segmentation in clothing.

In response to key challenges including the complexity of marine environments, the unique optical properties of water bodies, difficulties in data collection, and the lack of structured information in monocular optical images, this paper summarizes and analyzes recent significant advancements in equipment and theoretical methods related to 3D reconstruction of underwater monocular optical images. It dissects the existing issues in 3D reconstruction from three aspects: the need for enhanced maturity of monocular optical devices, insufficient research on the physical properties of underwater monocular optics, and the need for improved edge precision in reconstruction results. Addressing these key issues, we developed an underwater optical range-gated camera device and further studied a monocular optical image 3D reconstruction algorithm based on deep learning. This paper proposes potential key technologies for future 3D reconstruction of underwater scenes using monocular optical images.

The overexposure of laser image leads to poor imaging quality. A self-repairing method of overexposure of laser image under visual communication technology based on full-scale feature aggregation is proposed under visual communication. The overexposure feature points of laser image are detected by the deep fusion segmentation method of spectrum and texture features, and the differences of feature information are analyzed. The full-scale feature extraction of laser image is realized by the dynamic fusion method of visual communication combined with the changing regional distribution of pixels. The deep learning model is used to realize the visual aggregation of features, and the weights are allocated in the channel and spatial dimensions according to the differential characteristics of full-scale feature distribution. The self-healing of laser images is realized according to the view sensing and visual communication on multiple scales. Simulation tests show that compared with the other two methods, the PSNR value, SSIM value and FSIM value of the proposed method are higher, which are 38.56, 0.97 and 0.97 respectively, indicating that the overall quality, structure and details of the image repaired by the proposed method and the image feature retention are better. It shows that this method can reconstruct fine laser image by overexposure repair process, and effectively solve the problem of spatial information dispersion and spatial grid distribution imbalance caused by overexposure.

Accurate segmentation of 3D point cloud indoor plane elements is the basis of automatic reconstruction of indoor models. Aiming at the problem that the point cloud of boundary region is easily segmented incorrectly when the existing region growth algorithm is segmented indoor plane, a plane segmentation method with boundary feature constraint is proposed. This method first uses Euclidean clustering method to fuse RGB information to cluster the components that do not participate in plane segmentation, such as indoor tables and chairs, and then performs plane segmentation on the remaining point cloud. First, the region growth algorithm is used to segment the internal points in the plane, and then the growth process of seed points is monitored to accurately identify the boundary points in the neighborhood of seed points in the boundary region. The plane points are segmented by boundary points as growth constraint limits. Two sets of point cloud data in different scenes were used for experimental analysis. The test results show that the segmentation algorithm can accurately classify the point cloud in the boundary region, avoiding the problem of over -segmentation and under-segmentation of the point cloud. The accuracy and integrity of plane segmentation are improved by about 3% and 4% respectively compared with the region growth algorithm. The proposed algorithm can effectively improve the segmentation accuracy of indoor scenes.

When the laser beam encounters different media or interfaces, scattering phenomena occur, making the laser energy distribution uneven, resulting in blurred edges and loss of details in the dual wave scattering laser image. Therefore, an improved multi-scale Retinex based dual wave scattering laser image enhancement method is proposed. Perform downsampling on dual wave scattering laser images to obtain estimated blur kernel parameters. Update the light point parameters based on the fuzzy kernel, estimate the original clear image, and achieve deblurring processing of dual wave scattering laser images. Expand global enhancement on the dual wave scattering laser image to enhance the global contrast of the dual wave scattering laser image. Replace the Gaussian filter in the multi-scale Retinex algorithm with a guided filter to improve the multi-scale Retinex algorithm. Using an improved multi-scale Retinex algorithm to filter and enhance dual wave scattering laser images. Finally, the DWT image fusion algorithm is used to fuse the global contrast enhancement results with the filtering enhancement results, completing the enhancement of dual wave scattering laser images. The experimental results demonstrate that the proposed method has a good dual wave scattering laser image enhancement effect and can effectively improve the visual quality of the image.

The enhancement of details in complex graphic elements is affected by lighting changes, resulting in a decrease in contrast and clarity of graphic details observed by the human eye. Therefore, a method for enhancing the details of complex graphic elements in lighting changes based on laser vision communication is proposed. This method uses a multi-scale retina algorithm with color recovery function to unfold color correction on the collected graphics, and converts the graphics from the RGB color space to the HIS color space that is convenient for human observation through laser vision communication method. Adopting adaptive compensation method and limited contrast adaptive histogram equalization algorithm, the saturation I value and brightness S value of the graph are enhanced, and the enhanced image is converted back to the RGB color space. The experimental results show that this method has good color correction effect and strong stability in enhancing graphic details.

Laser active imaging technology plays an important role in many fields. However, due to the interference of various factors during the imaging process, images may generate noise, which affects subsequent information extraction and processing. Therefore, a laser active imaging visual image denoising method based on wavelet neural network is designed. Design a super-resolution reconstruction method based on image registration algorithm, which concentrates multiple frames of spot images in each subregion to correct the differences between images. Given that the color images captured by laser imaging are rich in primary colors, this leads to a huge amount of data and efficiency bottlenecks in processing. In order to optimize the subsequent preprocessing and recognition process, the average method is used to implement grayscale processing of the image. Design a wavelet neural network structure with a single hidden layer structure, with only one node set in the input layer to receive input information, and only one node set in the output layer to output processed results. Determine the number of hidden layer nodes according to the design method, and take the number of samples as the value of the number of image pixels used for learning, to achieve denoising processing of laser active imaging visual images. The experimental test results show that the denoised image of the design method is relatively clear while retaining image details. The difference in range index is small, and the pixel distribution of the denoised image is relatively uniform.

A hierarchical segmentation method for stereo images based on 3D laser scanning is proposed to address the current issue of poor hierarchical segmentation performance in stereo images. Firstly, based on a 3D laser scanner, collect 3D laser point cloud data from stereo images; Secondly, calibrate the correspondence between stereo image pixels and 3D laser point cloud data; Finally, hierarchical segmentation of stereo images is achieved through OTSU multi threshold segmentation method. The experimental results show that the average intersection to union ratio of the proposed method is between 83.6% and 92.5%, and the hierarchical segmentation accuracy of stereo images is high, the effect is good, and it is more suitable for practical applications.

Accurate prediction of communication link reliability is the key to ensure the selection of laser communication link. Therefore, the confidence interval prediction method of laser communication link reliability is studied to improve the availability and quality of service of communication link. The signal-to-noise ratio of the laser communication link is the reliability characteristic of the laser communication link; Through wavelet decomposition algorithm, the signal-to-noise ratio sequence of laser communication link is decomposed, and the noise sequence and stationary sequence are obtained; In the extreme learning machine, the stationary sequence is input to predict the stationary sequence of the signal-to-noise ratio of the laser communication link at the next moment; The standard deviation sequence of the noise sequence is solved and input into the extreme learning machine to predict the noise sequence of the signal-to-noise ratio of the laser communication link at the next time; According to the cumulative distribution function of Gaussian distribution, combined with the prediction results of the next time stationary sequence and noise sequence, the confidence interval of laser communication link reliability is determined, and the confidence interval prediction of laser communication link reliability is completed. Experiments show that the noise sequence obtained by this method is mainly concentrated in the range of ± 6 dbm, and the maximum error between the upper and lower bounds of the prediction results and the actual confidence interval is only 1.1 sbm, which is practical.

Multiple fiber cores in elastic optical networks have the function of transmitting information, but the distance between the fiber cores is small, which makes it easy to experience crosstalk and increases the error rate of elastic optical networks. To improve the service quality of elastic optical networks and effectively balance the allocation of network resources, a distributed machine learning based elastic optical network resource balance allocation method is proposed. Minimizing the maximum occupied frequency slot number as the optimization objective, establishing a crosstalk optimization model, and introducing the Pisces algorithm in distributed machine learning to solve the model, in order to improve the transmission quality of the fiber core. The resource allocation method of spectrum slicing is used to allocate resources in elastic optical networks, calculate the resource fragmentation rate of links in the network, allocate spectrum time resource windows for services based on the calculation results, and complete resource balancing allocation. The experimental results show that the proposed method has the highest resource utilization rate of 94%, the highest blocking rate of only 2.97%, and the highest energy consumption per unit bit of only 10 nJ/bit, which is practical.

In the Internet of Things, the state recognition of laser communication equipment is crucial for the accuracy of data transmission and scheduling. Currently, the device status mainly relies on the controllers and sensors in the DCS equipment, which are determined by setting a single threshold. But as the complexity of the equipment increases, the accuracy of this method is affected. To this end, a machine learning based method for laser communication equipment state classification and recognition was studied. Using time series sliding window mode to partition the state feature vectors of laser communication equipment; Define the abnormal status level of laser communication equipment based on its characteristic attributes that have an alarm effect; Building a laser communication equipment state recognition model based on machine learning fusion of alarm features. The experimental results show that by using different types of laser communication equipment as test objects and setting their fault states in different scenarios, the research method can achieve state recognition of test equipment in various scenarios, which has practical value.

The complex environment of wireless laser communication poses a serious challenge to the reliability of wireless laser communication signals. Therefore, a method for detecting abnormal wireless laser communication signals under multiplicative noise interference is proposed to take timely processing measures and ensure the quality of wireless laser communication. Integrating convolutional neural networks and machine learning methods, a wireless laser communication signal anomaly recognition model based on deep learning technology is constructed to determine the existence of anomalies. Based on the identified abnormal characteristics of wireless laser communication signals, estimate abnormal parameters such as center frequency point, pulse period, and scanning rate to complete wireless laser communication signal anomaly detection. The experimental results show that the normalized recognition indices for different signal anomalies are as high as 0.98, 1, 1, 0.99, and 0.99, and the normalized root mean square error is low, which proves that the proposed method has high detection accuracy and superior detection performance.

In order to improve the transmission characteristics of fiber-optic communication, a new air-hole arrangement structure ring photonic crystal fiber is designed, which has a large-diameter air-hole as the core, and the cladding is made of SiO2 as the base material, containing three layers of air-holes in the form of a rectangular, circular, and ring-cut-sector polygon. The transmission characteristics of photon orbital angular momentum modes in the designed fiber are analyzed based on the finite element method. The results show that the stable transmission of 118 OAM modes can be achieved in the high refractive index ring around the fiber core in the C+L band without the existence of radial higher-order modes, the maximum effective mode field area are 305.3 m2, which obtained in 1.63 m, which effectively increases its communication capacity, and the minimum effective refractive index difference between its neighboring vectorial modes is 5.43×10-3, the maximum confinement loss is 1.06×10-8 dB/m, the minimum nonlinear coefficient is 0.285 w-1 ·km-1, and the minimum dispersion of the vectorial modes is 2.68 ps/(km·nm), and the quality of intrinsic modes are all between 95.56%~98.37%, which could increase the quality of communication. This photonic crystal fiber can be applied to mode division multiplexing system, which provides a reference for improving the communication transmission efficiency.

Analyze the impact of atmospheric turbulence and atmospheric scintillation on the performance of laser communication, in order to maximize the quality of laser communication. Based on the constructed laser communication channel transmission model under atmospheric turbulence, this paper analyzes the atmospheric scintillation phenomenon caused by the uneven distribution of laser communication beam refractive index caused by atmospheric turbulence, which leads to an increase in laser beam intensity distribution. The atmospheric scintillation index is calculated, and combined with the probability density function of the instantaneous signal-to-noise ratio received by the laser communication system, the impact of atmospheric turbulence and atmospheric scintillation on the probability of laser communication interruption is analyzed The impact of average bit error rate and average channel capacity; The experimental results show that reducing transmission distance and zenith angle, increasing receiver aperture and average signal-to-noise ratio can reduce the impact of atmospheric scintillation on laser communication performance during atmospheric turbulence; The atmospheric scintillation index of low order laser beams changes more significantly when facing strong atmospheric turbulence, and the quality of laser communication can be improved by improving the order of laser beams; Moreover, OAM beams can better resist the impact of atmospheric turbulence and scintillation on laser communication performance compared to Gaussian beams.

In order to reduce the crack defects of the cladding layer of broadband laser solid forming, this study takes the minimum crack density as the optimization goal, through respectively response surface method design and GA -BP neural network model of crack defects optimized by genetic algorithm, the crack defect prediction model and the combination of process parameters with the least crack density are obtained. Compare the optimization results of the two. The results show that the crack density obtained by the response surface method is 0.075 mm/mm2, the crack degree obtained by the test under this parameter is 0.077 486 mm/mm2, and the relative error is 3.21%. The minimum crack density obtained by optimizing the process parameters of GA-BP neural network model is 0.057 2 mm/mm2, the experimental value is 0.058 123 mm/mm2, and the relative error is 1.59%. The effectiveness of parameter optimization of GA-BP neural network in the actual laser solid forming process is verified, which provides a theoretical basis for effectively eliminating or reducing crack defects.

In order to improve the localization accuracy of SLAM task of UAV in weak texture scene, a visual VO system based on point-line feature fusion, PLK2-SLAM, is proposed based on ORB-SLAM3 system. In this system, according to the statistical analysis of feature similarity, the features with large differences are retained to achieve the de-noising of features and improve the quality of features. In the feature matching process, the two-thread point feature matching method is used, the optical flow method is used for tracking features, and the LK2LBF method designed in this paper is used for feature reextraction, so as to reduce the calculation amount. A vector-based reprojection error method for line features is proposed to improve the accuracy of pose estimation. Finally, the performance of the PLK2 -SLAM system is verified on the open data set TUM. In the weak texture scene, the accuracy is improved by 84% compared with the classic point-and-line system PL-SLAM, which improves the stability and robustness of the ORB-SLAM3 system.

Currently, conventional methods for detecting internal defects in bearings are mainly based on enhancing the laser image of bearings so as to amplify the features of defects and realize defect detection by means of connectivity domain marking, etc. The detection accuracy is poor due to the low dimensionality of the filtering process of the image. In this regard, a laser detection method for internal bearing defects based on edge extraction and curve fitting is proposed. Firstly, the filter size is selected in combination with the filtering requirements, and according to the complexity of the image, one-dimensional as well as two-dimensional filtering is carried out respectively. Then the coordinates of the pixel points on the middle line of the binary distribution map of the derivative sign are extracted to mark the center line of the original diffraction stripe image. Finally, the defect edge line is extracted by combining the Sobel operator, and the edge line fitting process is realized by calculating the square and the mean difference of the defect edge, so that the background image and the defect contour image are separated from the original image to realize the defect detection. In the experiments, the detection accuracy of the proposed method is tested. The final test results show that when the proposed method is used to detect bearing defects, the peak signal-to-noise ratio and image structure similarity of the detected images is high, and the detection effect is more ideal.

Based on the rapid expansion of the production scale of the photovoltaic industry, this paper discusses the lack of accuracy in photovoltaic power prediction, the lack of tracking efficiency of maximum power points, and the backwardness of operation and maintenance technology of photovoltaic power plants. Combined with the artificial neural network (ANN) model, meta-heuristic algorithm and photovoltaic health state architecture technology, the internal mechanism of digital-real integration driving the high-quality development of the photovoltaic industry is deeply discussed. Multilayer Perceptron (MLP), Recurrent Neural Network (RNN), Long Short-Term Memory Neural Network (LSTM), Gated Recurrent Unit Network (GRU) and Convolutional Neural Network (CNN) were used to optimize the existing technology and improve the accuracy of PV power prediction. The meta-heuristic algorithm is introduced to help solve the problems of fast response, continuous oscillation at the maximum power point and easy locking of local peak points, and improve the tracking efficiency of the maximum power point. Data Quality Routines (DQRs), digital twins, and AI-driven fault diagnosis algorithms are used to solve the problem of lack of accurate, general, and location -independent data-driven PV diagnostic algorithms in the field of PV system fault detection, and improve the operation and maintenance procedures of large-scale PV power plants.

This study aims to prepare an iron-based alloy laser cladding coating with excellent wear resistance under high temperature and high speed conditions. By selecting steel based material metal discs as the basic material for cladding coatings, using iron powder as the main component, combined with high-purity chromium carbide powder and tungsten disulfide, and utilizing the auxiliary effect of laser cladding machines, iron-based alloy laser cladding coatings are prepared. The wear resistance of the coating was tested using a testing machine in a high-temperature and high-speed friction environment. The test results showed that the friction coefficient and wear rate of the coating reached 0.3 and 20% respectively at 400 ℃, belonging to a relatively high range. Moreover, at this temperature, the coating became smoother after friction and did not exhibit more severe particle wear. Under the 1500r/min test conditions, the friction coefficient of the coating showed a low state, and there were no obvious pits or scratches on the coating. From this, it can be seen that the coating has high wear resistance in high-temperature and high-speed environments.

The conventional intelligent human-computer interaction system mainly uses visual models to obtain user interaction positioning information, but it has certain positioning errors, resulting in a small number of system tasks. Therefore, the design of an intelligent human-computer interaction system based on laser vision sensors is proposed. In the hardware design of the system, design a full range laser vision sensor with dual joints for pitch and rotation. In the software design of the system, distortion coefficients are added to the positioning information returned by sensors through the conversion between the world coordinate system and the camera coordinate system, correcting the interaction positioning of system users, and based on the number of interaction instructions given by users, interaction instruction task constraints are planned, and intention recognition technology is introduced to analyze the characteristics of interaction instructions and perform corresponding operations, achieving the human-machine interaction function of the system. The system performance test results show that the system completes a larger number of tasks under the number of multi-level interactive instructions, and the system runs better.

Artificial mountains have rich levels of landscape hierarchy and are an important component of Chinese classical garden art. In modern garden design, they mainly play a role in increasing the local landscape effect. Due to the large volume of artificial mountain construction, it will carry a significant load and experience significant settlement deformation during long-term application. Once the deformation is too large, it can cause the mountain to crack. Therefore, a settlement monitoring method for plant garden rockeries based on 3D laser scanning technology is proposed. Determine the scanning density through the principle of 3D laser scanning, and obtain the 3D coordinates of the point cloud data of plant and garden rockeries according to standard parameters and scanning density; Using the collinear condition equation to decompose the three-dimensional coordinates of the point cloud, determine the relationship between the three-dimensional coordinates of the point cloud data and the spatial coordinates of the object, and obtain the inner and outer orientation elements in the point cloud coordinates; By using the spatial rendezvous calibration method and considering optical distortion, the orientation elements are calibrated to construct a coordinate difference equation for the main point of the image, obtain the settlement deformation of the plant garden rockery, and complete the monitoring method design. Taking a forest garden in a certain city as the test object, the research method was used to monitor the settlement of artificial rockeries in the garden. After data comparison, it was found that the research method has precise monitoring effects and can effectively complete the monitoring of rockery settlement deformation.

Non-contact 3D laser imaging technology is used to accurately detect 3D printing surface forming defects, and an automatic detection method of 3D printing surface forming defects based on 3D laser imaging technology is proposed. 3D laser point cloud data of 3D printed surface forming materials were obtained by using the scanner of 3D laser imaging system with vertical incidence scanning. The sample consistency initial registration algorithm is selected to carry out the coarse registration of laser point cloud. The accurate registration of laser point cloud is achieved by removing outliers and minimizing errors by using point-to-surface ICP algorithm. Using the four-quadrant near point search algorithm, linear interpolation is used to process the near 3D projection points of the defect pixels, and the 3D coordinate values of the defect points of 3D printing surface molding are obtained. The experimental results show that the method can effectively detect different types of defects such as holes and printing dislocation of 3D printing surface forming materials, and obtain accurate defect location results.

Previous infrared laser imaging target tracking methods have poor tracking performance due to only extracting a single feature of the infrared laser imaging target. Therefore, an infrared laser imaging target tracking method based on artificial intelligence technology was designed. By using infrared laser imaging equipment, a large amount of target data is obtained and corrected to construct an infrared laser imaging target model. With the help of artificial intelligence technology, multiple features such as shape, edge, and texture of the infrared laser imaging target are extracted, and feature similarity is calculated. The extracted features are matched with candidate features, and the mean shift of the features is calculated, Determine the position of the infrared laser imaging target. Through the above design, complete the design of the infrared laser imaging target tracking method. The experimental test results show that the tracking robustness of the proposed method can reach 99.11%, the EAO can reach 99%, and the infrared laser imaging target tracking effect is good.