摘要:Addressing the challenge of point ahead angle (PAA) in satellite-based free-space optical communication systems, this paper introduces a novel wavefront sensing method, the Projected Pupil Plane Pattern (PPPP), utilizing the transport-of-intensity equation (TIE). Laboratory experiments confirm its viability. The PPPP method, rooted in TIE, can deduce wavefront distortions due to atmospheric turbulence by analyzing variations in light intensity distribution over different transmission distances. Utilizing the back Rayleigh scattering from the communication laser, PPPP's atmospheric turbulence measurements align with the satellite's direction, offering an effective solution to PAA-related issues in satellite-ground laser communications. In our experiments, a 1-m ground telescope simulates an upward laser transmitter and captures the backscattered light for imaging. We measure wavefront distortions caused by atmospheric turbulence up to 10 km using backscattered light from altitudes of 10 km and 17 km. These distortions are simulated using a spatial light modulator or a transparent plastic sheet. The results demonstrate that PPPP and the commonly used Shack-Hartman wavefront sensor provide comparable wavefront reconstructions for various distortions, with the reconstructed phase's residual difference around 30% of the initial phase.
摘要:Addressing the need for precise non-contact measurement of semiconductor wafer thickness, this study introduces a method based on laser confocal technology that ensures remarkable accuracy. It utilizes a voice coil nanodisplacement platform for high-resolution actuation of a laser confocal optical probe, enabling precise axial scanning. This method relies on identifying the peak points on the confocal laser's axial response curve, which are indicative of the objective lens's focal point, to accurately align and position the wafer's upper and lower surfaces. By accurately calculating the physical coordinates of each sampling point on the wafer surface through ray tracing algorithms, this technique achieves high-precision non-contact measurement of wafer thickness. A specialized laser confocal sensor for semiconductor wafer thickness measurement was developed, showcasing an axial resolution of under 5 nm, an axial scanning range of up to 5.7 mm, and repeatability in thickness measurement of under 100 nm across six wafer types. The process takes less than 400 ms for a single wafer. This research successfully applies confocal focusing technology to semiconductor measurement, offering a novel solution for high-precision, non-destructive, online wafer thickness measurement.
摘要:To enhance the detection of alkane gas concentrations across a broad spectrum, this study introduces a mid-infrared telemetry and remote sensing approach utilizing the Cassegrain system. The design of the optical system's transmitting, receiving, and detecting modules was explored. Utilizing a 3 464 nm mid-infrared light source in conjunction with the Cassegrain system, an optical system for alkane concentration telemetry and remote sensing was developed. To accommodate target measurements within a 25-100 m range, an optical telemetry system was crafted with a 25 cm diameter. To counteract optical signal loss in the Cassegrain system, a transmission aspheric collimation structure was employed, enabling target focusing at varying distances. To address issues arising from the system's large aperture, a co-axial transmission and reception strategy, reducing disparity were implemented. The system can accurately measure targets within the 25-100 m range. Both primary and secondary mirrors were coated with a mid-infrared enhanced gold film, significantly boosting system stability. Testing reveals the system's overall transmission efficiency at 86%, with a reception efficiency of 75.8%. This alkane telemetry and remote sensing system facilitates remote gas concentration measurement over extensive ranges and distances, offering convenience, rapid measurement, and enhanced field safety. It ensures the optical system's stability and operational safety.
关键词:telemetry and remote sensing;alkane concentration measurement;mid-infrared light;Cassegrain system
摘要:Considering the critical need for a high-precision, six-degree-of-freedom measurement system in China's advanced manufacturing sector, this study introduces a novel approach for large-scale, high-accuracy attitude measurement utilizing a monocular vision module for laser tracking. Initially, the paper outlines the hardware configuration of the laser tracking measurement system, encompassing both the attitude measurement system and the cooperative target, and establishes a mathematical model to define spatial attitude angles accurately. It then delves into the adaptive clear imaging features of the attitude measurement module. By leveraging an optical distortion model and Zhengyou Zhang's calibration algorithm, it crafts a real-time camera imaging model that dynamically adjusts the pixel coordinates of feature points, enhancing the precision of feature point extraction. This is further refined by integrating the geometric traits of the cooperative target with the EPnP and SoftPOSIT algorithms, leading to a sophisticated attitude measurement technique complemented by an automatic monitoring and error correction mechanism. This dual approach enables precise automatic measurement across any attitude within the specified distance and measurement range. The system's accuracy was validated through experiments using a two-dimensional precision turntable equipped with a cooperative target, showing an impressive attitude measurement accuracy within a 3-10 m range, with yaw and pitch angles within ±30° and roll angle within ±180°. Specifically, accuracy was better than 0.049° with a 14-feature-point cooperative target and better than 0.065° with a 10-feature-point target. These findings underscore the superiority of this method over other recent laser tracking measurement techniques, highlighting its broad applicability and minimal constraints on the cooperative target's feature point configuration, thus fulfilling the precision measurement demands of high-end manufacturing's laser tracking and measurement.
摘要:In wireless communication, controlling electromagnetic wave propagation and polarization direction is crucial for signal detection and reception. This study introduces a liquid metal-based electromagnetic metasurface for polarization conversion in the X (8-12 GHz) and Ku (12-18 GHz) bands. It offers benefits like ultra-wideband performance, high polarization conversion ratio, compactness, durability against mechanical fatigue, flexibility, and affordability. The metasurface consists of a periodic arrangement with a top step-shaped liquid metal structure, a middle dielectric layer, and a bottom copper foil, capable of ultra-wideband cross-polarization conversion or wideband circular polarization conversion from 7.595 GHz to 17.712 GHz. With a 1.6 mm wide step-shaped liquid metal structure, it achieves over 90% polarization conversion ratio across 7.595-17.712 GHz with a 79.9% relative bandwidth, enabling co-polarization to cross-polarization conversion. At a 0.3mm structure width, it converts linear to circular polarization between 10.864-12.288 GHz with a 12.30% relative bandwidth. Additionally, in the 7.328-7.592 GHz band with a 3.54% relative bandwidth, it maintains over 90% polarization conversion ratio, facilitating co-to cross-polarization conversion. Experimental sample tests reveal a 4.20% relative error between experimental and simulated results, indicating strong theoretical and experimental correlation, thus confirming the metasurface's versatility and effectiveness.
摘要:To address the vibration challenges in flexible thin plate structures like solar panels on spacecraft, this study investigates a translational flexible hinged plate system. A binocular vision-based measurement and control experimental platform is developed. This platform employs the binocular stereo vision technique for vibration detection, and introduces a self-recurrent wavelet neural network controller (SRWNNC) to mitigate vibration. The system's binocular vision is precisely calibrated. Utilizing the principles of disparity and advanced image processing algorithms, it calculates the three-dimensional coordinates of specific markers to capture vibration signals. A finite element model of the system is constructed, facilitating the identification of system model parameters. Following this, the SRWNNC is trained within a simulation environment using the identified model parameters, aiming for effective vibration control in the experimental system. Experiments and simulations are conducted on the system, focusing on both fixed base and translational trajectory movements, to evaluate the effectiveness of binocular vision in vibration detection and the SRWNNC in active vibration suppression. The findings confirm that the binocular vision sensor achieves a high accuracy less than 0.1 mm in detecting vibrations, and the SRWNNC outperforms traditional large gain PD controllers in damping vibrations, thus validating the efficiency and accuracy of the proposed vibration detection and suppression methods.
摘要:To achieve swift, precise, and non-destructive detection of ball screw raceway's normal section, a measuring system based on an optical micrometer is developed. Initially, the detection device is crafted following the principle of measuring the ball screw raceway's normal section. The original data processing algorithm for the raceway's normal section is then refined, introducing the angle division arc method and arc data homogenization technique. Subsequently, potential errors from sensor installation, the horizontal moving platform's straightness, and sensor inaccuracies within the detection device are examined and compensated for. The detection system's accuracy, comprising the device and the enhanced algorithm, is validated through experimental tests. These tests reveal that with the algorithm's optimization, the normal section parameters of the raceway measured by the system show improved convergence as the divided arc angle increases. Post-optimization, the maximum measured values for arc radius and contact angle, along with the manufacturer's error margins, are 3.6 μm and 19′31″, respectively, marking enhancement rates of 64.36% and 53.46% over pre-optimization. Furthermore, the maximum standard deviation values for the measured arc radius and contact angle are 1.56 μm and 2′41″, respectively, with enhancement rates of 22.77% and 56.67%. Class A uncertainty for measured arc radius and contact angle dropped to 0.63 μm and 1′6″, respectively, a reduction of 56% and 70% compared to before the algorithm's optimization. The detection system's accuracy and repeatability for the raceway's normal cross-section meet the measurement standards.
关键词:optical micrometer;ball screw;raceway normal section;angle division arc method;arc data homogenization method
摘要:To solve the problem of missing absolute heading data and attitude drift in the visual-inertial navigation system, and to enhance its positioning accuracy, an adaptive visual-inertial-geomagnetic tightly integrated positioning system was developed for environments with unknown magnetic fields. Initially, the calibration process for the internal and external parameters of standard tri-axis magnetometers is detailed. Following this, a strategy for generating global and frame-to-frame constrained residuals from geomagnetic data is outlined. The system dynamically adjusts fusion weights based on variations in magnetic intensity and employs a nonlinear optimization approach for the visual-inertial-geomagnetic integration to estimate its motion state accurately. Outdoor tests conducted on a university campus demonstrated that the system remains stable amidst magnetic disturbances from buildings and vehicles, achieving positioning accuracy better than 0.8% (RMSE). When compared to VINS, this system reduces position error by an average of 24%, showcasing impressive real-time capabilities. Incorporating magnetometers and adaptive fusion techniques significantly boosts the performance of existing visual-inertial navigation systems, offering reliable real-time positioning for autonomous systems.
摘要:To address the challenge of target tracking for non-maneuvering, single-station setups in long-range scenarios, we propose a target tracking algorithm leveraging three-dimensional angle of arrival data, characterized by its asymptotically unbiased nature. Initially, we construct a motion and observation model centered on a non-maneuvering single station, assuming a known rate prior, and examine the system's observability. To tackle the bias inherent in the pseudo linear least squares algorithm, we introduce a constrained total least squares method that demonstrates asymptotically unbiased properties, with its effectiveness validated through simulations. In tests involving three-dimensional angle tracking over distances in the hundred-kilometer range, with angle measurement standard deviations at 0.1°, 0.2°, and 0.3°, the constrained total least squares method achieves a time-average relative distance error of 6%, 12%, and 21% within 50-100 seconds, respectively, and an absolute position error of 9 km, 19 km, and 35 km; at initial distances of 70, 140, and 280 km, the errors are 1%, 6%, and 30% for the same duration, with absolute errors of 0.7 km, 9 km, and 30 km. Notably, the relative distance error can be reduced to below 10% within 100 seconds, marking a significant precision enhancement, while maintaining operational speed comparable to the pseudo linear least squares method. The constrained total least squares approach exhibits rapid convergence, high accuracy, and swift processing, showing resilience against angle measurement errors and initial distance variations. It offers a robust solution for 3D angle of arrival tracking of non-maneuvering single-station targets in distant settings.
关键词:target tracking;3d arrival of angle;pseudo-linear;asymptotic unbiased estimation
摘要:Traditional FCM algorithms cluster based on raw data, risking distortion from noise, outliers, or other disruptions, which can degrade clustering outcomes. To bolster FCM's resilience, this study introduces a fuzzy C-means clustering algorithm that leverages adaptive neighbor information. This concept hinges on the similarity between data points, treating each point as a potential neighbor to others, albeit with varying degrees of similarity. By integrating the neighbor information of sample points, labeled GX, and that of cluster centers, labeled GV, into the standard FCM framework, the algorithm gains additional insights into data structure. This aids in steering the clustering process and enhances the algorithm's robustness. Three iterative methods are presented to implement this enhanced clustering model. When compared to leading clustering techniques, our approach demonstrates over a 10% improvement in clustering efficacy on select benchmark datasets. It undergoes thorough evaluation across different dimensions, including parameter sensitivity, convergence rate, and through ablation studies, confirming its practicality and efficiency.
摘要:To address the challenges of color distortion, blurred details, and low contrast in underwater images caused by complex imaging environments, a novel restoration method that integrates color compensation with dual background light fusion is introduced. This method begins by enhancing the traditional underwater imaging model to reflect light absorption and attenuation in water more accurately. It incorporates a color compensation technique inspired by Retinex theory and white balance algorithms to mitigate the impact of the water's background color. A novel dual-candidate background light fusion approach is then developed to precisely estimate the global background light, considering the intensity and color distribution of background light. This is followed by the use of guided high-pass filtering to refine and boost the transmission across each channel, leveraging the connection between the water's background color and the scattering coefficient, without relying on specific water environment parameters. The method concludes with the restoration of the underwater image through the reverse application of the imaging model. Testing on four diverse underwater datasets has shown that this approach surpasses several classic and advanced methods, delivering images with more natural colors and enhanced, clearer textures. The color difference value improved by 5.4%, while the UCIQE and FDUM metrics increase by 8.3% and 4.5% respectively, underscoring the method's effectiveness in both qualitative and quantitative evaluations and its significant contribution to enhancing the quality of underwater imagery.
摘要:To address the issue of low recognition accuracy in lightweight algorithms for steel surface defect detection, this paper introduces a Multi-scale Enhanced Feature Fusion (EFF) technique. Initially, an Adaptive Weighted Fusion (AWF) module calculates fusion weights adaptively for different feature levels. This allows shallow features to enrich with deep semantics without compromising detail. Subsequently, the Spatial Feature Enhancement (SFE) module boosts the fused features from three distinct directions and improves network stability by integrating residual pathways, enabling the convolution process to extract more critical information. The model then selects better training samples based on the overlap between the prior box and the ground truth. Experimental outcomes show that the proposed method achieves a detection accuracy of 80.47%, marking a 6.81% increase over the baseline algorithm. Moreover, with 2.36 M parameters and 952.67 MFLOPs, this algorithm efficiently and accurately identifies steel surface defects, demonstrating significant practical utility.
摘要:To tackle the challenges in multimodal classification tasks involving hyperspectral images (HSI) and LiDAR data, such as cross-modal information expression and feature alignment, this paper introduces a contrastive learning-based multi-branch CNN-Transformer network (CLCT-Net) for the joint classification of hyperspectral and LiDAR data. Initially, CLCT-Net employs a feature extraction module with a ConvNeXt V2 Block to capture shared features across different modalities, addressing the semantic alignment issue between data from heterogeneous sensors. It then develops a dual-branch HSI encoder with spatial channel and spectral context branches, alongside a LiDAR encoder enhanced by a frequency domain self-attention mechanism, to secure more comprehensive feature representations. Lastly, it leverages ensemble contrastive learning for classification to further refine the accuracy of multimodal collaborative classification. Experimental evaluations on the Houston 2013 and Trento datasets demonstrate that the proposed model excels in extracting and integrating cross-modal data features, achieving superior ground object classification accuracies of 92.01% and 98.90%, respectively, when compared to existing models for classifying hyperspectral images and LiDAR data.
摘要:Under low-light conditions, the fusion of infrared and visible images often results in images with poor contrast, lacking in detail, and requiring a lengthy processing time. To address these issues, this paper introduces an enhanced multi-scale structural fusion approach. Initially, it improves the low-light visible image using a dynamic range compression enhancement algorithm. Subsequently, through a multi-scale structural image decomposition method, it separates the enhanced visible and infrared images into their low-frequency and high-frequency components. For image fusion, the low-frequency components of both image types are merged using a technique based on the root mean square error coefficient. The high-frequency components are initially fused in a straightforward manner, followed by an optimized fusion using a self-adaptive weight adjustment based on image information entropy. Afterward, by reversing the multi-scale structural decomposition, the fused low and high-frequency components are combined to form a complete image. To further enhance the image contrast, a regional pixel enhancement algorithm based on gray level classification is introduced. The effectiveness of this method is compared with nine conventional infrared and visible image fusion techniques, both qualitatively and quantitatively, using TNO and CVC-14 datasets. The proposed method demonstrates superior performance in terms of average gradient, cross entropy, edge intensity, standard deviation, and spatial frequency, along with an improved overall visual quality. This confirms that the images produced by the proposed method exhibit enhanced detail, clarity, contrast, and are processed more quickly.
关键词:image processing;multi-scale structural fusion;dynamic range compression;root mean square error;information entropy;Contrast