White-light interferometry is widely employed in precision manufacturing sectors, including semiconductor processing, thin-film thickness measurement, and metal defect detection, owing to its merits of high precision, high measurement speed, and non-contact nature. The frequency domain analysis (FDA) technique has been applied to surface profile measurement by utilizing the slope of the phase distribution of the white-light scanning interference signal in the wavenumber domain. Through simulations and experiments, this study demonstrates that the accuracy of the phase obtained in the wavenumber domain is related to the amplitude distribution, whether in the presence of additive noise or phase noise, and quantitatively provides the optimal phase selection range for the FDA method. When the sample has strong spectral absorption, parts of the amplitude distribution in the wavenumber domain decrease significantly, rendering the corresponding phase information invalid and unreliable. To address this limitation, an improved algorithm is proposed in this paper, which incorporates the wavenumber-domain amplitude as a weighting factor into the least-squares fitting procedure. This method avoids the selection of the phase range for the FDA method and ensures high measurement accuracy.

In the simulation, the zero-mean Gaussian noise with a standard deviation (STD) of 10 nm is used as the phase noise ε_n(z) to generate a white-light interference signal. The Fourier transform is performed on the signal to obtain the amplitude I_F($\sigma $) and phase ($\sigma $) in the wavenumber domain. For the FDA method, the selected phase range $R_\sigma $ is defined as the wavenumber range corresponding to amplitudes in I_F($\sigma $) that are greater than CImax, where the coefficient C∈(0, 1) and I_max is the maximum value of I_F($\sigma $). The STD of z_s represents the measurement accuracy of the FDA approach. The influence of phase noise on the measurement accuracy and its relationship with amplitude distribution I_F($\sigma $) are investigated by calculating the STD of z_s across different $R_\sigma $. Furthermore, additive noise is added to the white-light interference signal with phase noise, and the STD of z_s is evaluated under varying $R_\sigma $ to determine the optimal phase selection range. Since conventional phase selection is not suitable for interference signals with complicated spectral distributions, this paper proposes a special frequency domain algorithm based on amplitude optimization in the wavenumber domain (amplitude optimization for frequency domain analysis, AOFDA). In this improved strategy, the amplitude distribution I_F($\sigma $) is introduced as a weighting factor in the phase linear fitting process to obtain the optimized surface measurement parameter z_w.

The simulation results indicate that when only phase noise exists, the accuracy of phase extraction improves with the increase in the amplitude corresponding to the selected phase, thereby enhancing the overall measurement precision. However, the phase extraction accuracy has an upper limit. When the range $R_\sigma $ corresponding to C≥0.5 is selected in the FDA method, the measurement results exhibit higher accuracy. Additive noise with different signal-to-noise ratio (SNR) is introduced into the white-light interference signal containing phase noise. For interference signals with the same SNR, the STD of z_s obtained with C≥0.5 is stable and relatively low. For signals with stronger additive noise, the STD of z_s shows a slight increase when C>0.5. In summary, if only phase noise exists, selecting the range $R_\sigma $ corresponding to C≥0.5 yields high measurement precision. When both phase noise and additive noise exist, C=0.5 provides higher measurement precision. Therefore, the $R_\sigma $ corresponding to C=0.5 is the optimal phase selection range for the FDA method.

To verify the feasibility of the AOFDA method for interference signals with complex spectral distributions, a band-stop filter spectral distribution was used as the absorption spectrum of the measured object in the simulation. In the band-stop range, simulations with different reflectivities were conducted to investigate the measurement accuracy. When the reflectivity of the sample is ≥20%, it exerts little influence on the measurement accuracy of the FDA method. However, the reflectivity of ＜20% moves the measurement results away from their true values. When the reflectivity of the sample is ＜20%, the measurement results obtained by the AOFDA method always exhibit better performance. This indicates that the measurement accuracy of the AOFDA method is largely unaffected by the absorption spectrum of the sample.

The flat surface of wedged window and the step surface were measured in the experiment. When the different value of C is selected to obtain the surface profile of the wedged window, the C=0.5 has the better performance that agrees with simulation result. The S_z of the wedged window surface calculated by the AOFDA method and the FDA method (C=0.5) are 34.6 nm and 50.2 nm, and the S_q are 5.5 nm and 6.7 nm, respectively. The average step heights, whose value is regarded as 3.0 μm, of 10 measurements calculated by the AOFDA method and the FDA method (C=0.5) are 3.0054 μm and 2.9566 μm, and the repeatability values are 17.9 nm and 19.4 nm, respectively. The experimental results demonstrate that the AOFDA method can achieve the measurement accuracy comparable to that of the FDA method when selecting the optimal wavenumber-domain phase range. Since the selection of an appropriate wavenumber-domain phase range is no longer required for the AOFDA method, it enables the white-light interferometer to better adapt to complex measurement environments.

The measurement accuracy of the FDA method is degraded by environmental vibration and background noise. This work confirms that the measurement precision of the FDA method is correlated with the amplitude distribution in the wavenumber domain through the selection of diverse phase ranges, and proposes a feasible phase selection strategy to enhance its measurement performance. It is verified that the FDA method is difficult to adapt to the situation where the sample has a strong absorption spectrum. To address these limitations, the AOFDA method is developed in this paper, which incorporates the wavenumber-domain amplitude distribution as a weighting function into the phase fitting procedure. The AOFDA method can achieve a measurement accuracy comparable to that of the FDA method processed with the optimal phase range, while eliminating the need for determination of the optimal phase range.

Achieving both a large field of view and high resolution remains a persistent and critical bottleneck in the field of optical imaging. Conventional imaging systems are constrained by a fundamental trade-off: expanding the field of view typically leads to a noticeable reduction in spatial resolution or requires the adoption of optical components that are bulkier, heavier, and more complex in structure, which limits their applicability in scenarios requiring portability or miniaturization. Existing solutions aimed at addressing this dilemma, such as those relying on mechanical translation stages, oscillating mirrors, or multi-lens arrays, suffer from inherent limitations. Mechanical translation stages are restricted by mechanical inertia, resulting in scanning frequencies as low as 10 Hz that severely hinder imaging efficiency; oscillating mirrors are prone to mirror surface deformation due to centrifugal forces at high speeds, with their maximum scanning frequency capped at around 100 Hz; while multi-lens array systems significantly increase overall system complexity, require highly efficient processing algorithms, and rely excessively on precise calibration procedures, with an equivalent scanning frequency of merely 50 Hz. To overcome these challenges, this thesis conducts in-depth research on large field of view stitching imaging technology based on galvanometer fast scanning. The primary objective is to fully leverage the high-speed scanning advantage of galvanometers (capable of kHz-level scanning frequencies, far exceeding traditional mechanical systems) to construct an efficient and reliable image acquisition system with an overlap compensation mechanism. Simultaneously, the traditional Harris algorithm, a classic corner detection method in computer vision, is improved to enhance its performance in feature extraction and matching for stitching tasks. Ultimately, this research develops a practical method for acquiring overlapping sub-images and seamlessly stitching them into a single high-resolution panoramic image, thereby resolving the trade-off between field of view and resolution.

The research methodology is based on two core aspects: the construction of a scanning imaging system with an overlap compensation mechanism and the introduction of key improvements to the traditional Harris corner detection algorithm. The scanning imaging system was constructed using a high-precision galvanometer and a GigE industrial camera, enabling the rapid acquisition of a grid of sub-images. On the algorithmic front, the thesis proposes two main modifications. First, a priori non-extremely large value suppression shape is adopted. This method determines the optimal suppression shape (square, circle, or ellipse) by calculating gradient directions from downsampled images, allowing the algorithm to adapt to different image textures. Second, a center region shielding adaptive non-extremely large value suppression method based on the overlap rate is proposed. This technique detects the horizontal and vertical overlap ratios between images and strategically applies suppression only within the overlapping regions, effectively reducing interference from non-overlapping areas. These improvements enhance the accuracy of corner point screening and the computational efficiency of the algorithm. The overall stitching process follows a designed "group-of-four" iterative strategy and employs homography estimation and weighted linear blending for seamless image fusion.

The effectiveness of the proposed method was rigorously verified through simulation and experimental studies. Simulation verification involved partitioning ten high-resolution 9000 pixel×12000 pixel source images into sub-arrays with preset overlap rates of 30%, 40%, 50%, and 60%. The results demonstrated that, compared with the traditional adaptive non-extremely large value suppression algorithm, the proposed method achieved notably faster computational speed. It also generally yielded a higher number of valid corner matches and a higher correct matching rate, with the advantage being particularly pronounced at lower overlap rates. Computational efficiency was improved by more than 60% under the different overlapping rates tested. The stitched image exhibited high quality, with PSNR values exceeding 40 dB and SSIM values close to 1, indicating excellent fidelity and minimal distortion. For experimental validation, a 6×11 array of 2592 pixel×1944 pixel images was first captured using the constructed galvanometer scanning system. Applying the improved algorithm to this dataset successfully generated a seamless panoramic image with a final resolution of 9553 pixel×19149 pixel. Compared to the traditional method, the complete stitching process achieved a 39.69% improvement in computational efficiency and a 10.31% increase in the feature matching correct rate, which directly contributed to significantly enhanced visual consistency in the final large-field-of-view composite. These results substantiate that the integration of high-speed galvanometer scanning with the enhanced Harris algorithm effectively addresses key limitations such as the low scanning speed of conventional mechanical systems and the computational inefficiency of traditional global feature detection, establishing a practical and effective approach for high-resolution imaging over a large field of view.

This research successfully develops a large field of view stitching imaging technology based on galvanometer scanning. By constructing a rapid scanning system and improving the Harris algorithm through a priori suppression shape and overlap-rate-based center shielding adaptive non-extremely large value suppression, the study provides an effective solution to the classic trade-off between field of view and resolution. The technology demonstrates substantial improvements in processing speed and stitching accuracy, enabling the generation of ultra-high-resolution panoramic images. The results offer a new technical paradigm for applications ranging from astronomical observation and satellite remote sensing to pathology slice scanning and endoscopic imaging. Future work may involve integrating deep learning techniques to further optimize the algorithm's adaptability and efficiency in complex scenarios.

Industrial steel surface defect detection is a core step of industrial quality inspection. However, traditional detection algorithms suffer from low accuracy, high model complexity, and difficult deployment due to the various types and large scale variations of steel surface defects. Existing deep learning-based detection algorithms have improved detection accuracy to a certain extent, but most of them feature large model size and high computational complexity, which makes them hard to deploy on resource-constrained industrial edge devices and thus restricts their practical application in real-time industrial detection scenarios. This study aims to propose a lightweight steel surface defect detection algorithm that balances high detection accuracy and low computational complexity, realizes efficient feature extraction for multi-scale steel surface defects, and meets the requirements of real-time detection and edge device deployment in industrial scenarios.

A lightweight steel surface defect detection algorithm named RGCE-YOLO was proposed based on an improved YOLOv11n. The backbone network was reconstructed with a novel HGRCBlock structure combining lightweight HGNet and improved reparameterized RepLightConv, and multi-branch convolution and normalization steps were fused during inference to reduce computational load while maintaining feature extraction capability. For the neck network, the lightweight GSConv replaced the original downsampling structure to reduce redundant information via mixed convolution and address the accuracy bottleneck of depthwise separable convolution; the C3K2 structure was redesigned by integrating the Spatial and channel synergistic attention (SCSA) mechanism to alleviate sub-feature semantic differences and enhance multi-scale object feature fusion. The CMBC_Detect module, constructed based on EfficientNet's inverted bottleneck structure and combined with the coordinate attention (CA) mechanism, was introduced into the detection head’s regression branch; the SE attention was replaced with the CA mechanism to retain spatial detail information, enhance small object detection ability, and compress the structure simultaneously. Channel-wise knowledge distillation (CWD) was applied to the improved model: each channel’s feature map was normalized to a soft probability map, and the asymmetric KL divergence between the teacher and student models was minimized to make the student model focus on salient regions of defect features, reducing feature loss without increasing computational cost. A self-built steel surface defect dataset with four defect types (rust, scratch, pit, crack) was established and expanded via data augmentation strategies including random cropping, dynamic brightness adjustment, salt-and-pepper noise, and Gaussian noise, and then divided into training, validation, and test sets at an 8:1:1 ratio. Ablation experiments verified the effectiveness of each improved module, and comparative experiments on distillation methods and mainstream detection algorithms were conducted under the same conditions. Generalization experiments were carried out on the public NEU-DET and GC10-DET datasets, and the model was converted to ONNX format for deployment verification on an ARM industrial computer. Precision, mAP50, mAP50-95, parameters, FPS, and FLOPs were used as key evaluation indicators to comprehensively assess the algorithm's performance.

Ablation experiments confirmed the effective optimization role of each improved module: HGRCBlock reduced model parameters and FLOPs while enhancing multi-scale feature extraction; GSConv reduced the neck network’s redundant information and computational complexity; SCSA improved the model’s perception of multi-scale objects and detection robustness; CMBC_Detect strengthened the capture ability for small, easily missed defects and further compressed the model size; CWD improved model accuracy without changing model size and computational complexity. On the self-built dataset, RGCE-YOLO achieved a precision of 83.4% (3.9% higher than YOLOv11n), with mAP50 and mAP50-95 increased by 3.9% and 4.6% to 80.5% and 50.3%, respectively; parameters decreased by 34.5% to 1.69 M and FLOPs by 33.3% to 4.4 G, realizing dual optimization of accuracy and lightweight design. Distillation comparison experiments showed that CWD had a better optimization effect than other distillation methods, and YOLOv11m was selected as the optimal teacher model by balancing performance and complexity. Mainstream algorithm comparison experiments indicated that RGCE-YOLO outperformed traditional algorithms (Faster R-CNN, RetinaNet, SSD) and other lightweight algorithms (MobileNetV4, StarNet, NanoDet-Plus-m) in detection accuracy, and had obvious lightweight advantages over other YOLO series algorithms, with an FPS of 185.8 meeting industrial real-time detection needs. Generalization experiments showed that RGCE-YOLO achieved a precision of 81.2% (8.4% higher than YOLOv11n) on the NEU-DET dataset and 73.1% (4.2% higher than YOLOv11n) on the GC10-DET dataset, with stable mAP improvements, demonstrating excellent generalization ability. ONNX deployment verification on the ARM industrial computer showed that the model maintained a precision of 81.7% and an FPS of 35.1 with no significant performance decline, proving its adaptability to edge devices. Visual results showed that RGCE-YOLO had a lower missed detection rate, stronger small object feature extraction ability, and more accurate bounding box positioning than YOLOv11n. The excellent performance of RGCE-YOLO stemmed from the synergy of all improved modules: RepLightConv’s reparameterization balanced feature representation and computational efficiency; GSConv and SCSA solved feature redundancy and semantic differences in multi-scale defect detection; the CA mechanism in CMBC_Detect retained spatial details and solved small object detection information loss; CWD enhanced dense object feature capture and generalization without additional computational cost.

RGCE-YOLO effectively solves the problems of low accuracy and poor lightweight performance of existing steel surface defect detection algorithms, and the synergy of the improved modules achieves dual optimization of detection accuracy and model lightweight design. On the self-built dataset, the algorithm not only has significantly higher precision, mAP50, and mAP50-95 than the YOLOv11n benchmark model, but also greatly reduces parameters and computational complexity, with its inference speed meeting industrial real-time detection requirements. It exhibits excellent generalization performance on public steel surface defect datasets and can be stably deployed on edge devices with no obvious detection performance decline, providing a reliable technical solution for real-time industrial steel surface defect detection. The design idea of combining a lightweight network structure, an attention mechanism, and knowledge distillation in this study also provides a reference for the lightweight optimization of other industrial object detection algorithms. In the future, constructing a dataset with more steel surface defect types and conducting targeted improvements for tiny object defects will further enhance the model’s comprehensive detection ability, promoting its wider application in industrial quality inspection scenarios with more complex defect types and higher detection requirements.

Nighttime low-light imaging is often affected by underexposure, limited dynamic range, and amplified sensor noise. These factors weaken edge, texture, and contour cues, and they increase background interference and feature ambiguity. Such degradation reduces detector reliability and causes missed detections, false positives, and localization bias, with more severe impact on small objects, distant targets, and dim instances. Based on YOLOv11n, LMD-YOLO was proposed as a low-light object detection method that integrates multi-scale features and local detail enhancement. Without substantially increasing model complexity, it strengthens multi-scale feature fusion and local detail extraction for low-light images and reduces high noise sensitivity in low-illumination scenes, thereby improving low-light detection accuracy.

LMD-YOLO combined multi-scale feature fusion and local detail enhancement through four components. 1) A Multi-Pool Spatial Pyramid Fast Pooling module (MPSPPF) was inserted into the backbone. It contained parallel pooling branches and a gradient-enhanced concatenation part. The parallel pooling used MaxPool2d and AvgPool2d. Max pooling strengthened salient responses and benefited key-object extraction in low-contrast images. Average pooling produced smoother representations and reduced noise sensitivity. The pooled outputs were fused with weighted aggregation to preserve global structure while retaining local detail cues. MPSPPF also used repeated sequential pooling. The same pooling operation was applied stage by stage to the previous pooled output to form a deeper pyramid. This design enlarged the effective receptive field with limited cost and built a richer feature hierarchy. It captured local and global context that supported recognition of blurred, partially occluded, or weakly illuminated objects. The gradient-enhanced concatenation improved information flow across pooling stages and reduced detail attenuation in early feature extraction. 2) A Cross Stage Partial–Enhanced Dual Hybrid Attention Network module (CSP-EDHAN) was constructed for fine-grained low-light features. Depthwise separable convolution reduced computation while maintaining spatial sensitivity. Residual connections stabilized feature propagation under severe noise. A multi-path fusion structure increased cross-stage information flow and strengthened weak-target cues. A dual hybrid attention mechanism combined channel emphasis and spatial selection. Channel emphasis highlighted informative channels under illumination degradation. Spatial selection focused responses on object regions and suppressed random-noise activations. This module improved feature contrast for dim targets and reduced false activations caused by background clutter. 3) A Dynamic Detail–Semantic Fusion Pyramid Network (DDS-FPN) was designed for neck fusion. A channel–spatial attention unit from the SDFM module guided feature selection during cross-scale aggregation. Shallow detail cues and deep semantic cues were fused with dynamic weights, which raised responses in target regions and lowered responses in noisy background regions. An adaptive downsampling module (ADown) aligned deep semantics with shallow details during scale transitions. It reduced feature misalignment, improved feature consistency across pyramid levels, and compensated for detail loss caused by low illumination. A bidirectional pyramid path strengthened bottom-up detail propagation and top-down semantic guidance. Multi-scale features were aligned and enhanced for small, medium, and large objects under low-light conditions. 4) A dual-group detection head was proposed to balance accuracy and efficiency. Input multi-scale feature maps first passed through a shared feature-extraction stem composed of two grouped convolution layers. Grouped convolution increased feature diversity at low cost and improved localization and recognition. The classification and regression branches shared the stem parameters to reduce redundant computation. Two independent output branches were then applied. Each branch used a 1×1 convolution layer to predict class probabilities and bounding-box distributions. The decoupled branch design reduced task interference and improved training stability for noisy low-light inputs.

Comparative experiments were conducted on EXDark and PASCAL VOC 2012 with YOLOv11n as the baseline. On EXDark, LMD-YOLO improved precision, recall, mAP50, and mAP50-95 by 1.3%, 2.3%, 3.1%, and 2.2%. On PASCAL VOC 2012, the same metrics increased by 1.4%, 1.5%, 1.8%, and 1.3%. Additional evaluation on a complex low-light scene dataset showed gains of 2.2% in mAP50 and 1.3% in mAP50-95 over the baseline. The results indicated that MPSPPF reduced early noise sensitivity through multi-branch pooling and context aggregation. Weighted fusion preserved structure cues while limiting noise amplification. Sequential pooling improved long-range context modeling, which benefited targets with weak boundaries. CSP-EDHAN strengthened weak-target representation through cross-stage fusion and attention-guided enhancement. Depthwise separable convolution preserved efficiency and improved feature robustness. DDS-FPN improved multi-scale alignment through dynamic detail–semantic fusion and attention-based selection. ADown reduced information loss during downsampling and improved semantic–detail consistency across pyramid levels. The dual-group head maintained lightweight computation and improved optimization through decoupled branches. Performance gains on PASCAL VOC 2012 suggested improved general feature quality rather than low-light domain bias. The method also showed stable behavior under mixed illumination and cluttered backgrounds, which are common in real nighttime scenes.

LMD-YOLO improves low-light object detection through noise-robust feature extraction, weak-target enhancement, and dynamic multi-scale fusion while preserving a lightweight design. The method achieves consistent accuracy gains on low-light benchmarks and remains effective under normal illumination. Future work targets broader illumination distributions, stronger robustness to extreme noise, glare, and motion blur, and improved generalization in natural scenes.

Micro-vibrations generated by internal moving components during spacecraft on-orbit operations constitute a critical factor limiting the optical axis stability and imaging quality of space-based optical payloads. Isolating these disturbances and maintaining precise pointing are essential for high-precision tracking and imaging tasks. Incorporating a payload platform capable of both vibration isolation and precision pointing is considered a "multi-win" technical solution. While piezoelectric actuators are common, their limited stroke restricts millimeter-level applications. Voice coil motors (VCMs), conversely, offer large strokes and high linearity but often lack guidance in fast steering mirror applications, leading to lateral drift. Furthermore, traditional hexapod platforms based on the Gough-Stewart configuration, while accurate, involve six-degree-of-freedom fully coupled control. This introduces structural redundancy and control complexity for payloads that primarily require pitch and yaw stabilization. To address the dual requirements of vibration isolation and tracking pointing in pitch and yaw directions, a simplified quadruped platform integrated with voice coil actuation is proposed. The design targets a natural frequency of approximately 5 Hz to facilitate passive isolation of high-frequency disturbances and a pointing adjustment range of ±1° for low-frequency tracking. This study aims to balance structural simplification, large-stroke capability, and high-precision motion, providing a physical foundation for subsequent active vibration control and compound axis tracking systems.

A simplified multi-degree-of-freedom parallel mechanism configuration was adopted. The platform consists of a payload platform, a base, and four supporting legs arranged symmetrically at 90° intervals along a circumference. A push-pull actuation strategy was implemented where legs 1 and 3 control pitch motion, and legs 2 and 4 control yaw motion, enabling decoupled control. VCMs were selected as actuators to leverage their zero stiffness, large stroke, and high resolution. To ensure precise axial motion and eliminate friction, a flexible guidance structure and flexure hinges were developed. A double diaphragm spring system was designed to provide low axial stiffness for actuation compliance while maintaining high radial stiffness to prevent lateral drift and contact between the coil and magnet. The diaphragm springs were designed using 60Si2Mn steel, and Finite Element Analysis (FEA) was used to verify structural integrity. A theoretical model based on dynamic principles established the relationships between leg inclination, pointing range, and resonant frequencies. FEA was further conducted to validate the design, simulating the stiffness of the diaphragm springs and performing modal analysis on a complete 3D model with a 40 kg payload to identify vibration modes. A physical prototype was manufactured, and a specific suspension system utilizing a tripod, tension springs, and flexible cords was constructed. This system effectively offloaded the gravity of the 40 kg payload to simulate a micro-gravity environment, ensuring that the platform's dynamic characteristics were tested without the bias of static gravitational forces. An experimental test system comprising accelerometers, linear optical encoders, and a frequency response analyzer was established to perform sine sweep excitations (1 Hz to 300 Hz), stroke capability tests, and decoupling verification.

FEA results indicated that the double diaphragm spring configuration achieved an axial stiffness of 3.83 N/mm and a radial stiffness of 1703.87 N/mm. Stress analysis showed a maximum stress of 330.33 MPa at the spring's flexure points, which is well below the 1180 MPa yield strength of the 60Si2Mn material, resulting in a safety factor of 3.57. Simulation confirmed that a leg stroke of 2.7 mm was sufficient to achieve a platform tilt of 1°. Modal analysis predicted the first and second order modes (pitch and yaw) at 4.7413 Hz, aligning closely with the 5 Hz design target, while the third mode (Z-axis translation) appeared at 5.7381 Hz. Higher-order modes were pushed beyond 60 Hz, avoiding interference with the control bandwidth. Experimental testing on the prototype demonstrated the following: 1) Individual legs achieved a total stroke exceeding 6.8 mm. Under sinusoidal actuation, the platform successfully verified a pointing range of ±1°, meeting tracking requirements. 2) During single-axis excitation, parasitic motion in the non-driven axis was minimal. For a 2 mm pitch motion, yaw cross-coupling was approximately 2% (0.041 mm), confirming effective decoupling. 3) With a 40 kg payload, measured resonant frequencies were 5.2 Hz for pitch and 5.6 Hz for yaw. These values agree well with the FEA prediction (4.74 Hz). 4) Tests with varying loads validated the dynamic model: 20 kg load resulted in ~7 Hz resonances, and no-load conditions yielded ~16 Hz.

A voice coil-actuated integrated vibration isolation and pointing quadruped platform was successfully designed and verified. The simplified parallel mechanism, utilizing flexure hinges and double diaphragm springs, effectively resolves the conflict between large-stroke pointing and high-precision vibration isolation. Experimental results confirmed the platform meets critical specifications: a natural frequency near 5 Hz under a 40 kg load and a pointing range of ±1°. The study demonstrates that pitch and yaw motions are effectively decoupled, and the experimental data aligns well with theoretical models. This work establishes a verified physical platform capable of passive high-frequency isolation and active low-frequency tracking. Future developments may include integrating this platform with a fast steering mirror to form a two-stage composite control system, enabling coarse tracking via the platform and fine tracking via the mirror for enhanced precision.

Homodyne coherent wireless optical communication enables the direct recovery of baseband signals by interfering a local oscillator with the received signal light under identical frequency and stable phase conditions. This approach provides outstanding detection sensitivity, strong resilience against interference, and high spectral efficiency, making it a promising candidate for next generation high capacity and long reach optical links. However, its practical realization depends critically on maintaining precise and stable phase alignment between the signal and the local oscillator. This requirement imposes rigorous demands on laser source coherence, phase locking accuracy, and adaptive control algorithms. Addressing these challenges is fundamental to harnessing the full potential of homodyne detection in real world dynamic environments such as free space optical channels, which are often affected by atmospheric turbulence, polarization fluctuations, and other time varying impairments.

Significant progress has been made in the key technologies of the robust homodyne coherent receiver, including wavefront distortion correction technology, local oscillator noise suppression technology, phase locking technology, polarization control technology, mixing technology and dual balanced detection technology. Wavefront distortion correction, primarily implemented through adaptive optical systems, compensates for phase errors induced by atmospheric turbulence and optical imperfections. Early conceptual work proposed using deformable mirrors, and subsequent development of wavefront sensors enabled closed loop operation. Modern adaptive optical systems employ multi actuator deformable mirrors, liquid crystal spatial light modulators for static or quasi static correction, and hybrid schemes that combine fast steering mirrors with high order deformable mirrors. These improvements enhance correction bandwidth, spatial resolution, and suitability for real time operation in turbulent channels. Local oscillator noise suppression targets phase, frequency, intensity, and quantum noise sources. Optical injection locking and self injection locking techniques effectively reduce phase noise. Electronic feedback and servo control, often using PID algorithms, provide precise frequency stabilization. Intensity noise is mitigated through balanced detection architectures and by operating the local oscillator at an optimal power level that maximizes the signal to noise ratio. Quantum noise, a fundamental limit, is addressed through techniques such as squeezed state generation and nonlinear optical processing, though these remain areas of active research. Phase locking technology ensures the local oscillator accurately tracks the signal phase. Optical phase locked loops form the core of this subsystem, evolving from discrete component based designs to architectures incorporating digital signal processing. Key developments include decision driven optical phase locked loops, Costas loop based designs for carrier recovery, and hybrid optical phase locked loops that combine optical and electronic feedback. Phase recovery algorithms, adapted from digital coherent communications, play a complementary role. The Viterbi Viterbi algorithm, blind phase search, and Kalman filtering are widely used to estimate and compensate residual phase noise, especially for high order modulation formats like 16 QAM and 64 QAM. Polarization control ensures the polarization states of the signal and local oscillator are matched at the mixer. Early systems relied on mechanical polarization controllers with slow feedback. The field progressed towards integrated electro optic devices based on lithium niobate, offering faster response. Recent trends focus on system level integration, combining polarization tracking with phase recovery algorithms, and on novel devices like metasurface based polarization controllers that offer compact size and programmable functionality. Polarization diversity receivers and polarization insensitive mixer designs also help alleviate the control burden. Mixing techniques generate the in phase (I) and quadrature (Q) components required for complex signal recovery. Fiber based mixers using 3 dB couplers were common but suffer from polarization sensitivity. Integrated optical hybrids on platforms like silica or indium phosphide offer improved stability and compactness. Crystal based mixers, particularly those utilizing the electro optic effect in periodically poled lithium niobate, provide inherent phase accuracy and polarization insensitive operation. Recent advances in thin film lithium niobate enable broadband, low loss mixers suitable for high speed applications. Dual balanced detection converts the optical interference signals into differential electrical outputs, effectively rejecting common mode noise and amplifying the weak signal component. The core of this technology is a pair of matched photodiodes in a differential configuration. Developments have focused on improving photodetector bandwidth, responsivity matching, and integration with planar lightwave circuits. Systems achieving common mode rejection ratios exceeding 40 dB have been demonstrated, directly translating to higher receiver sensitivity and dynamic range.

Homodyne coherent systems exhibit superior sensitivity and spectral efficiency compared to conventional heterodyne and direct detection schemes. The continuous refinement of wavefront correction, noise suppression, phase locking, polarization control, mixing, and balanced detection has progressively lowered implementation barriers, enabling laboratory demonstrations and early field trials. Current performance benchmarks show clear advantages in receiver sensitivity, tolerance to optical signal to noise ratio, and resilience to certain channel impairments. Nevertheless, challenges remain for widespread deployment. Further reduction of phase noise is critical for higher order modulation formats. Managing the coupled dynamics of phase, polarization, and frequency under rapidly varying atmospheric conditions requires more intelligent, adaptive control systems. System complexity and cost also need reduction for practical applications. Future research will likely focus on several key directions: 1) Deep photonic integration, combining lasers, modulators, mixers, and detectors on a single chip using platforms like silicon photonics or thin film lithium niobate to reduce size, power consumption, and cost. 2) Advanced digital signal processing and machine learning algorithms for joint estimation and compensation of multiple impairments in real time. 3) Exploration of quantum inspired techniques, such as phase squeezed states, to surpass classical sensitivity limits. 4) Development of robust and adaptive adaptive optical systems tailored for specific turbulent channel characteristics. Progress along these paths will enhance the robustness, scalability, and energy efficiency of homodyne coherent receivers, solidifying their role in future high performance terrestrial and satellite based wireless optical communication networks.

In clinical practice, pathological examination is considered the "gold standard" for diagnosing various types of cancer. In the field of pathological image intelligent diagnosis, the detection performance of deep learning models typically relies on large-scale and high-quality manually annotated datasets. Due to the high cost and expertise required in annotating medical images, large-scale annotation is often infeasible, resulting in challenges such as limited training data and significant domain variation in clinical scenarios. Therefore, achieving accurate, efficient, and generalizable medical image recognition under limited annotation conditions has become a important challenge in the field of intelligent medical image processing. Mainstream methods adopt multiple instance learning (MIL), which enables whole slide image(WSI) classification using only slide-level labels. However, MIL-based approaches still suffer from issues such as overly concentrated attention allocation and model overfitting when handling complex pathological patterns.

To address these challenges, this study proposes a MIL framework based on a mixed distillation training strategy. First, Starting from publicly available datasets, the processes of image segmentation, image partitioning, and image feature extraction for digital pathology slides have been implemented. The OTSU segmentation method is used to remove the background images that do not contain any pathological information. Each WSI in the dataset is divided into non-overlapping image patchs to facilitate model training and feature extraction. These preprocessing steps help improve the efficiency and accuracy of the model, enabling it to focus more on learning the key pathological features. The self-supervised Vision Transformer-S/16(ViT-S/16) was used to train the model on unlabeled pathological image data. The ViT-S/16, which is a self-supervised learning method based on contrastive learning, can learn key feature representations from images. This process enables the model to learn transferable features from unlabeled data and provide better initialization for the downstream classification task. Then, a mixed distillation mechanism multiple instance learning (MDM-MIL) model integrating dual-path adaptive attention and random instance concealing is designed. The adaptive attention module dynamically adjusts the degree of focus on different texture features and captures semantic relationships between them to further enhance the classification accuracy. The first path adaptive gated attention submodule calculates the attention scores of all instances, and then aggregates the attention scores adjusted by random instance masking to integrate the remaining instance information. The second path uses an adaptive multi-head attention submodule to calculate the attention scores of all instances, and then aggregates the attention scores adjusted by random instance masking to integrate the remaining instance information. This adaptive method not only enhances the model's ability to recognize local features but also significantly improves its understanding of global context. Meanwhile, the random instance concealing (RIC) mechanism randomly masks instances with high attention scores during training, forcing the model to explore additional informative regions and mitigating overfitting to a few critical patches. Mixed distillation mechanism (MDM) integrates the feature information of the remaining instances from the two paths after being processed by the RIC method, thereby enhancing the learning efficiency and classification accuracy of the model. This method captures semantic relationships between different texture features of an image to further increase the classification accuracy. Finally, the classifier achieved accurate classification results on a dataset of breast cancer whole slide pathological images. This paper proposes a multi-instance learning model based on a mixed distillation strategy to address the issues of excessive attention allocation and model overfitting in existing multi instance learning. This model combines a collaborative training dual-path adaptive attention model with a random instance masking method, and ultimately uses a mixed distillation strategy to concatenate the aggregation results of the dual paths. The model can fully learn more key instance feature information while considering the correlation between various types of instances, ultimately improving the performance of the model and alleviating model overfitting.

This study has the characteristics of uneven distribution of categories and weakened sample labels. Therefore, AUC and F1-score are comprehensively adopted as the main evaluation indicators to ensure the scientificity and comprehensiveness of the performance assessment. Experimental results demonstrate the superiority of the proposed method. On the Camelyon16 dataset, the model achieves an accuracy of 98.445%, an AUC of 0.984, and an F1-score of 0.983. On the BRACS dataset, the corresponding values reach 77.852%, 0.905, and 0.764, respectively, showing significant improvements over existing MIL approaches. The experimental results on the Camelyon16 and BRACS datasets show that compared with other advanced multiple instance learning models, This model has significantly better classification metrics, indicating that our model has more advantages in pathological image classification.

This thesis proposes novel algorithm designs and conducts comparative experimental evaluations on corresponding public datasets, achieving superior results compared to the current state-of-the-art methods. This paper contributes to the advancement of automatic analysis techniques for pathological images based on deep learning, which holds significance in assisting cancer diagnosis, and improving the efficiency of clinical analysis.

Changes in retinal vascular morphology serve as crucial indicators for diagnosing various diseases, such as diabetic retinopathy, hypertension, and cataracts. Traditional diagnosis of retinal diseases primarily relies on manual assessment by clinicians, typically involving the analysis of retinal vascular curvature, width, and branching patterns based on imaging results from fundus examinations, optical coherence tomography, and fluorescein angiography. However, these imaging outcomes are susceptible to interference from factors such as illumination and noise, which increase the difficulty for clinicians in distinguishing retinal vessels from the background. Furthermore, inherent ocular structures like the optic cup and optic disc regions, as well as pathological alterations such as hard exudates, soft exudates, and the macular area, can also impede clinicians' judgment. Diagnostic results often vary subjectively among different physicians, potentially leading to misdiagnosis due to insufficient experience or fatigue. Automated retinal vessel segmentation technology can not only significantly reduce the time and effort required for manual annotation by clinicians and improve diagnostic efficiency but also minimize the impact of human factors to the greatest extent, providing more objective results to assist physicians in decision-making. Therefore, developing an accurate and automated retinal vessel segmentation technique holds substantial practical significance.

In the field of retinal vessel segmentation, existing methods can be broadly categorized into unsupervised and supervised approaches. Unsupervised methods do not rely on manually annotated data and primarily achieve segmentation by analyzing the morphological features of vessels. Common techniques include matched filtering-based, edge detection-based, and model-based methods. Matched filtering methods mainly utilize Gaussian kernel functions or their variants, which resemble the cross-sectional shape of blood vessels, to construct two-dimensional linear filter banks. These methods enhance tubular structures in retinal images through multi-scale and multi-directional filter responses, combined with thresholding for preliminary detection of vessel pixels. Edge detection-based methods capture boundary features between vessels and the background by calculating gradient changes in pixel intensity or using differential operators such as Canny and Sobel, often integrated with local or global thresholding strategies to extract vascular contours with continuous topological structures. Model-based methods rely on geometric or statistical models, such as active contour models, level set methods, or vascular tubular models, to mathematically describe vessel structure and distribution. They iteratively optimize energy functionals or probabilistic fitting processes to gradually evolve fine segmentation boundaries that adaptively conform to vessel morphology. However, such methods often depend on manually designed feature extraction pipelines, making it difficult to fully capture the complex topological structures of vessels, thereby resulting in limited segmentation performance. In contrast, supervised methods require expert-annotated data as training samples to achieve higher accuracy in vessel segmentation tasks. These methods can be further divided into traditional machine learning methods and deep learning methods. Traditional machine learning methods typically involve two steps: manual feature extraction and classification. Handcrafted features mainly include edge responses, wavelet responses, principal component information, and color intensity information. Based on the choice of classifier, they can be roughly categorized into k-nearest neighbors, random forests, support vector machines, and Gaussian mixture models. Traditional machine learning methods extract segmentation targets based on vascular features and fundus image characteristics. Compared to manual screening, they offer advantages such as shorter processing time and faster segmentation. However, their feature extraction process still requires manual design and cannot adequately represent the multi-scale structural features of vessels. With the continuous advancement and development of deep learning, it has gradually been adopted across various tasks in computer vision. In the field of medical image analysis, segmentation algorithms have been continuously updated and iterated. The introduction of various deep learning models and modules has further improved the accuracy and efficiency of fundus vessel segmentation algorithms. Compared to traditional machine learning methods, deep learning methods can automatically learn optimal features and achieve predictive modeling without human intervention, offering higher segmentation efficiency. Nevertheless, challenges such as limited representation of fine vessels, interference from complex backgrounds, and discontinuities in vessel structures remain. To address the aforementioned issues, this paper proposes a retinal vessel segmentation algorithm based on Grouped Channel Transformer and Modulated Fusion. First, a dual-dimensional cooperative attention module is designed in the encoder to enhance the distinguishability of vessel features and improve the model's accuracy in segmenting microvessels. Second, a Grouped Channel Transformer module is applied to construct a dual-encoder architecture, establishing global dependencies to effectively capture detailed vessel information. Finally, a Modulated Fusion module is introduced at the decoder to enhance feature fusion between the encoder and decoder, thereby improving the model's robustness against noise interference.

Experiments conducted on the DRIVE, CHASE_DB1, and STARE datasets demonstrate that the proposed algorithm achieves accuracy rates of 0.9711, 0.9762, and 0.9758, sensitivities of 0.8026, 0.8219, and 0.8042, and specificities of 0.9873, 0.9866, and 0.9899, respectively. The proposed method outperforms other algorithms in overall segmentation performance, demonstrating superior segmentation capability.

GTMF-Net demonstrates robust vessel segmentation capabilities on the DRIVE, CHASE_DB1, and STARE datasets. It achieves a favorable balance between model complexity and segmentation performance, offering both relatively low computational cost and high segmentation accuracy. However, the algorithm currently exhibits limitations in feature extraction within low-contrast regions, which can lead to mis-segmentation. Future work will focus on optimizing the algorithm's design to improve the accuracy of vessel identification in pathological areas.

Remote sensing image target detection plays a crucial role in a wide range of practical applications, including resource exploration, urban planning and management, environmental monitoring, and disaster emergency response. With the rapid development of high-resolution sensors and large-scale data acquisition technologies, the demand for accurate and efficient object detection methods has increased significantly. However, targets in remote sensing images are typically characterized by small size, large scale variations, complex and cluttered backgrounds, blurred boundaries, and insufficient texture information, which significantly increases the difficulty of effective feature extraction and reliable detection. In addition, factors such as imaging altitude, atmospheric interference, and illumination variations further degrade image quality, resulting in low contrast between targets and the surrounding environment. Despite substantial progress in recent years, small target detection in remote sensing imagery still faces several critical challenges. From a localization perspective, commonly used metrics such as Intersection over Union are overly sensitive to slight variations in bounding boxes when applied to small targets, leading to unstable regression optimization and reduced localization accuracy. From a data perspective, the limited number of small target samples and the imbalance in category distribution often cause models to underrepresent small object features, resulting in missed detections. Moreover, in complex environments, small targets are often mixed with background regions that share similar textures or visual patterns, making it more difficult to accurately separate foreground from background and achieve precise localization.Therefore, improving the accuracy, robustness, and generalization capability of small target detection in remote sensing images, while enhancing training stability and localization precision under complex conditions, remains a critical and meaningful research problem.

Addressing the above challenges, a remote sensing small target detection algorithm integrating attention mechanisms and feature enhancement is proposed. First, a multi-level feature collaboration strategy is constructed by combining channel attention and spatial attention with a gating mechanism. This design enables adaptive weighting of features across different levels, strengthens cross-scale information interaction, and effectively suppresses background interference. By dynamically emphasizing salient regions and reducing redundant responses, the model is able to capture subtle and discriminative features of small targets more reliably. Second, a multi-scale feature enhancement module is designed by integrating three-dimensional convolution with multi-scale feature encoding to form an enhanced neck structure. This module facilitates the fusion of semantic and spatial information across different scales, allowing the network to preserve global contextual information while retaining local details. The use of three-dimensional convolution further enhances the modeling of inter-scale dependencies, improving the representation of small targets under complex backgrounds and varying resolutions. Finally, SIoU and NWD are jointly employed as regression loss functions to optimize bounding box prediction. This combination improves geometric alignment between predicted and ground-truth boxes, reduces sensitivity to scale differences, and stabilizes the optimization process. As a result, the model achieves faster convergence and more accurate localization, especially for small and densely distributed targets.

Extensive experiments are conducted on three widely used remote sensing datasets, including NWPU VHR-10, DIOR, and RSOD, to evaluate the effectiveness and generalization capability of the proposed method. The results show that, compared with the baseline model, the improved algorithm achieves consistent performance gains across all datasets, with mAP@0.5 increasing by 9.9%, 3.1%, and 3.2%, respectively. These improvements demonstrate that the proposed method effectively enhances feature representation and detection performance for small targets. Visualization results further indicate that the model can successfully identify targets that are easily missed by conventional approaches, particularly in scenarios with dense distributions and complex backgrounds. In addition, the proposed method shows improved robustness in distinguishing targets from visually similar background regions. Ablation studies confirm the effectiveness of each component, showing that the multi-level feature collaboration strategy, the multi-scale feature enhancement module, and the optimized loss function contribute synergistically to performance improvement. The results also suggest that the proposed framework maintains stable detection performance under different dataset distributions, indicating good generalization ability.

In this study, an attention- and feature-enhancement-based small target detection algorithm for remote sensing images is proposed. By introducing a multi-level feature collaboration strategy and a multi-scale feature enhancement module, the method significantly improves cross-level feature interaction and strengthens the representation of small targets. Meanwhile, the joint optimization of SIoU and NWD loss functions further enhances localization accuracy and training efficiency. The proposed approach effectively addresses key challenges in remote sensing small target detection, including complex backgrounds, scale variations, and insufficient feature information. Experimental results on multiple benchmark datasets confirm the superiority, robustness, and generalization capability of the proposed method over the baseline model. These findings demonstrate its strong potential for practical applications in real-world remote sensing tasks, particularly in scenarios that require accurate detection of small and densely distributed targets. Furthermore, the proposed framework provides a reliable technical reference for future research and can serve as a foundation for further improvements in small target detection methods.

Salient Object Detection (SOD) is difficult in low-light conditions. The images often suffer from poor quality, heavy noise in dark areas, and low contrast. These issues make it hard for models to detect objects accurately and reliably. Current methods mostly rely on single-mode features. These methods cannot capture complex information in low-light images. This is especially true when light is uneven or local details are lost. Therefore, this study proposes an efficient and robust hybrid model. We combine the strengths of Convolutional Neural Networks (CNNs) and Transformers. Our goal is to improve detection performance and overcome the limits of traditional methods in the dark.

This study proposes a Dual-Stream Hybrid Network (DSHNet) based on a hybrid CNN-Transformer architecture, which adopts an asymmetric dual-branch encoder-decoder lightweight structure to make up for the defects of the existing hybrid models in low-light image processing, such as inflexible multi-scale image processing and low feature fusion efficiency. DSHNet combines two main strengths: first, CNNs extract local features at multiple scales with shallow convolution suppressing noise interference caused by uneven illumination and deep convolution capturing fine-grained texture features; second, Transformers provide powerful global modeling with self-attention mechanism breaking the bottleneck of modeling long-distance dependence between target pixels in low-light images. The model uses a multi-level strategy to extract features from both the CNN and Transformer modules, then adopts a global-local synergy method to realize cross-hierarchical bidirectional fusion of features from different branches at the same scale, solving the problem of information loss in multi-scale feature fusion. We also designed a Channel Spatial Attention Enhancement Module (CSAEM), which dynamically fuses channel-spatial attention and channel recalibration, and uses a 1×1 convolution layer to learn the weighted coefficient to adaptively balance channel and spatial attention, realizing explicit perception of edge structures and effectively suppressing background noise in low-light images. In the encoding stage, the Efficient Spatial Reduction Attention (ESRA) module is applied step by step to Transformer features, compressing the keys and values in multi-head self-attention through convolution operations and introducing the DropKey mechanism to reduce the number of model parameters and computational complexity while avoiding over-focus on local areas. We also adopt an overlapping patch embedding strategy and embed depth convolution in the feedforward network to enhance the continuity of local features and avoid the dilution effect of fixed position encoding on dark area features. In the decoding stage, we use a two-stage hierarchical design with a coarse-grained decoding branch restoring the spatial resolution of global semantic features and a fine-grained decoding branch refining the dark texture and local contrast, and deploy saliency prediction heads at six feature stages with a deep supervision strategy to realize gradient backpropagation from coarse to fine and strengthen the boundary constraints of dark area targets. Finally, we use a combined loss function of Focal Loss and Dice Loss. This helps the model focus on difficult samples and improves area overlap, effectively solving the issue of unclear boundaries caused by severe imbalance between foreground and background in low-light scenes.

To verify the effectiveness of DSHNet, we constructed a complex low-light scene mixed dataset LlSOD, which includes 3960 images and corresponding binary masks by screening and mixing LLVIP and YLLSOD datasets with NIQE greater than 6.78 and labeling with Labelme. Experimental validation was conducted on LlSOD and standard datasets including RGB252, RGBD-385, RGBT-621 and VI-789, with experimental configuration set as 50 training epochs, initial learning rate of 5e-5, poly learning rate strategy with power=0.9, weight decay of 1e-4 and batch size of 4. Ablation experiments on the LlSOD dataset show that the introduction of the lightweight CNN branch only increases 0.3M parameters, while the F-measure is increased by 5.2% and MAE is reduced by 2.4% compared with the single Transformer branch; the CSAEM dual-branch enhancement module further reduces MAE to 0.093 and increases F-measure to 0.901, which verifies that the synergy of channel and spatial attention is superior to a single attention mechanism. Comparative experiments show that DSHNet is much lighter than the similar hybrid model ABiUNet, reducing the number of parameters by 23% (only 26.1MB). On the LlSOD dataset, it achieved high scores: 91.2% for F-measure, 83.7% for S-measure, and 87.2% for weighted F-measure, which are 5.4%, 5.7% and 4.8% higher than the YLLSOD model dedicated to low-light image SOD respectively. In the generalization performance test on standard datasets, DSHNet achieved 0.021 MAE and 0.852 F-measure on RGB252, 0.037 MAE and 0.907 F-measure on VI-789, and even showed competitive performance on dual-modal datasets RGBD-385 and RGBT-621 without introducing additional modal data, with 0.850 F-measure on RGBD-385 and 0.854 F-measure on RGBT-621, which proves that the global-local feature fusion of DSHNet can simulate multi-modal information interaction and has strong robustness. Visual comparison results show that DSHNet can stably realize object positioning and segmentation in various low-light scenes such as backlight, indoor night and outdoor night, effectively solving the problems of foreground and background segmentation difficulty and inaccurate object positioning faced by most methods in low-light environments.

In conclusion, DSHNet successfully solves key problems in low-light object detection such as difficult balance between global modeling ability and computational efficiency, difficulty in considering global information and local detail features, and overlap between multi-scale feature fusion and low-light noise. It combines the advantages of both CNNs and Transformers, and realizes the lightweight design of the model on the premise of improving detection performance. The results prove that our model is both lightweight and powerful, and provides a new effective approach for single-modal low-light image saliency detection. In the future, we will continue to optimize the model structure, further reduce the computational complexity of the model while maintaining detection accuracy, and plan to explore more real-world applications such as autonomous driving and security monitoring to promote the engineering transformation and application of this technology.

Phase measurement profilometry (PMP) has emerged as a key optical technique for acquiring three-dimensional (3D) surface information. It has been successfully adopted in fields such as precision manufacturing, biomedical imaging, robotics, security inspection, and digital entertainment. Recently, the demand for high-resolution optical 3D measurement has increased rapidly. In the high-resolution PMP system with digital projector, the measurement error caused by pixel matching between projector and camera becomes obvious. In ultra-high-resolution PMP systems with a high projector-to-camera pixel size ratio (PPSR), fringe patterns projected by diamond pixel digital light processing (DLP) projectors exhibit prominent zigzag edges. Furthermore, the sampling frequency of vertical fringes is only half that of horizontal fringes, resulting in a significant decline in phase measurement accuracy and becoming a key bottleneck restricting system performance. Addressing this critical issue requires a comprehensive solution that mitigates both pixel dimension induced and discrete sampling errors.

A high-precision measurement method combining pixel remapping and deep learning was proposed to overcome these challenges. Firstly, by analyzing the transmission characteristics of fringes in the PMP system and the pixel mapping law of diamond pixel projectors, a pixel remapping scheme for column-interlaced diamond linear fringes (CIDLF) was designed. This scheme rearranges diamond pixels into column-interleaved straight-line arrangements, doubling the sampling frequency of vertical fringes while effectively alleviating the zigzag phenomenon on fringe edges. Unlike the traditional diamond zigzagged fringe (DZF) and rectangle linear fringe (RLF), CIDLF maintains consistent fringe width between design and projection, ensuring better sinusoidality of projected fringes. Secondly, a ResUNet deep neural network integrating residual network and UNet architectures was constructed to smooth the captured fringes. The UNet component enables spatial downsampling and upsampling for high-level feature extraction and detail recovery through skip connections, which fuse low-level pixel information with high-level semantic features. Residual modules incorporated into both encoder and decoder paths alleviate gradient vanishing and overfitting during deep network training, enhancing the network’s ability to preserve fine object details while smoothing fringes. The network’s basic convolution block (BasicConv) consists of 3×3 convolution, batch normalization, dropout regularization, and rectified linear unit activation, ensuring efficient feature learning and noise suppression. To ensure reliable network training, a high-quality dataset construction method combining defocused projection and the 32-step phase-shifting algorithm was developed. The focused fringes and defocused fringes corresponding to the focused one are captured. Then, parameters of fringes including background, amplitude and phase are obtained with 32-step phase-shifting algorithm and the averages of the background and amplitude of the two types of fringes are worked out. A high contrast smooth fringe is generated using the above parameters as the ground truth. The dataset included 19200 fringe images for training, 400 for validation, and 400 for testing, covering fringe patterns with periods of 8, 12, and 16 pixels. The ResUNet was trained using stochastic gradient descent (SGD) optimization with a mean squared error (MSE) loss function. Training spanned 100 epochs with a batch size of 8; the initial learning rate of 5×10⁻¹ was updated to 5×10⁻² after 20 epochs, resulting in a final training MSE of 1.0128×10⁻⁵ and a validation MSE of 5.4261×10⁻⁵.

An experimental system composed of a LightCrafter 4500 projector (1140 pixel×912 pixel) and a Bonito CL-400C camera (2320 pixel×1736 pixel) with a Nikon zoom lens was established to verify the method. PPSR was adjusted by modifying the camera lens focal length, covering values from 0.5 to 85. Experimental results demonstrated that as PPSR increased, the pixel structure of the projector in fringe patterns became more distinct, leading to more prominent regular distributions of phase measurement errors. For traditional DZF, the phase error reached 0.3013 rad in the focused state, while CIDLF achieved a significantly lower error of 0.0898 rad, outperforming both DZF and RLF across all focusing conditions. When combined with ResUNet processing, the phase measurement error of CIDLF under the 4-step phase-shifting algorithm was reduced to 0.0225 rad, approaching the 0.0223 rad error of the 32-step algorithm. This performance is particularly valuable for high-speed 3D measurement applications that require fewer phase shifts. In measurements of step-shaped objects and objects with step reflectivity, CIDLF processed by ResUNet showed phase curves nearly identical to those of defocused CIDLF, with steeper edge transitions and smoother reflectivity boundary regions compared to processed DZF. For complex objects, the proposed method maintained low phase measurement errors even in intricate regions, outperforming defocused projection and DZF processing. The method effectively balances fringe smoothness and detail preservation, resolving the trade-off between error reduction and information retention in high-PPSR systems.

This method effectively resolves the measurement error issues caused by diamond pixel DLP projectors under high PPSR conditions. By combining pixel remapping to optimize fringe sampling and ResUNet to refine captured images, it achieves both high measurement accuracy and detail preservation. The proposed dataset construction method ensures robust network training, enabling the system to perform well with fewer phase shifts. This work provides a feasible approach for achieving high-precision, high-resolution, and high-speed 3D measurement. Future research will focus on adapting the network to varying system configurations and expanding its applicability to dynamic measurement scenarios.

As semiconductor manufacturing progresses towards sub-Xnm nodes and adopts 3D device architectures, such as gate-all-around field-effect transistors and 3D-NAND, wafer structures now feature extremely high aspect ratios and complex heterogeneous integration. This development imposes strict requirements on optical inspection and metrology, which are crucial for process control. Traditional inspection methods, which regard illumination as a static and passive source and mainly rely on backend algorithmic processing, are becoming increasingly insufficient. Their drawbacks, including instability in dynamic environments, substantial signal-to-noise issues in detecting nanoscale defects, and limited three-dimensional awareness, highlight the necessity for a revolutionary change. Therefore, next-generation optical inspection systems should transform into intelligent, integrated platforms based on the synergy of two key technologies: dynamic light field manipulation and multidimensional information fusion perception. This reconfigures the system from a passive observer to an active probe, where lighting and sensing are dynamically coupled to optimize detection from the physical signal generation stage.

The application of light field shaping in engineering clearly and logically evolved to cope with ever more challenging demands of semiconductor inspection. The first generation was called static light field shaping, addressing the fundamental need of converting an inherently Gaussian profile output from most lasers into useful, uniform profiles required by scanning applications. This was done using passive optical elements such as a diffractive optical element and a microlens array that were able to create a flat-top line illumination or flat-top spot illumination, thus establishing a fundamental hardware framework. With increasing complexity in inspection conditions and the need for flexibility, it evolved into active light field control: active, programmable devices, such as digital micromirror devices or liquid-crystal spatial light modulators. All of these contributed to a vital capability: the ability to shift quickly from one type of light to another, or to generate complex, programmable, structured light patterns in real-time and thus match the light field to particle inspection tasks. However, only achieving dynamic control did not ensure stable, high-fidelity performance for multiple tools and over time. This realization gave rise to step three: adaptive light field correction. We now looked inwards at compensating for the intrinsic flaws of our systems and their variations. With tools such as wavefront correctors, integrated beam pre-processing modules, and computational algorithms that enabled active measurement and correction of the static aberrations due to component tolerances or assembly stresses. This “ pre-standardization” of the input beam ensured that downstream modulation stages received a consistent, high-quality light field, which is critical for tool-to-tool matching and long-term stability. The ultimate stage in this progression is closed-loop light field self-calibration, which looks outward to combat real-time environmental perturbations. To build a high-bandwidth “ sense-analyze-act” feedback loop—using fast sensors and smart processors (usually augmented with artificial intelligence), and fast actuators—whereby the system has the capability of sensing and correcting in real time, transient disturbances such as vibration or thermal drift, transforming into an autonomous adaptive agent. Based on such a four-phase evolutionary process, with the principle of “ multi-dimensional fusion and collaborative control”, we develop a three-layer engineering application architecture for addressing the critical challenges in contemporary examination. It should be noted that this design can be viewed as a natural combination of these three levels: the physical execution layer forms its foundation, including devices such as spatial light modulators or digital micromirror devices, which directly shape the light field’s phase, amplitude, and pattern. The physical principal layer is the strategy layer that turns physical principles into practical strategies for breaking through conventional barriers. It applies techniques like structured illumination microscopy to obtain super-resolution images, active noise immunity and polarization engineering to improve the signal-to-noise ratio, and quantitative phase imaging for 3D topography sensing. Computational algorithm layer is an intelligent decision maker that drives the whole system by solving inverse problems, conducting AI-based defect classification, and allowing for adaptive optimization and closed-loop control, which ties everything together. There is a tight coupling between all three of those layers—with instructions guiding how they are implemented on hardware, and software refining both—that constitute an integrated smart system that is ready to tackle any challenging inspection task.

We have traced the evolution of light field control from a passive element into that which forms the heart of an active intelligent probing system for semiconductor optical inspection. Logical progression of the ideas, starting from static shaping and moving towards closed-loop self-calibration, paves the way forward towards realizing dynamic, resilient optical systems. Our suggested three-tiered application architecture provides an overall systematic approach for designing the next generation of inspection tools, effectively channeling advances in hardware, physics, and algorithms to overcome the intertwined problem of resolution, accuracy, speed, and three-dimensional vision. In the future, further development in this direction depends on closer integration with state-of-the-art computer intelligence and new optical materials. Highlights of the frontier include generative artificial intelligence driven inverse design that promises to discover non-intuitive, high-efficiency photonic architectures beyond human’s intuition; development of deep learning enhanced expert system, which is a hybridization between data driven learning and structured physics knowledge for robust and efficient optimization; and research on large model-based autonomous agent to perceive and make decisions about a whole scene in complicated inspection environment. Meanwhile, with the emergence of new hardware such as dynamic metasurface integrated photonic circuits, we may be able to make even more compact and flexible modulators for controlling light fields. We believe that this combination of smart software (algorithms) and hardware devices would enable autonomous and powerful light field manipulation, consolidating its position as the essential “eyes” for precision manufacturing of the semiconductor industry and beyond.