Integral imaging three-dimensional (II3D) display technology has achieved significant progress in applications such as 3D television, medical imaging, and virtual reality. However, II3D displays still face limitations, such as low resolution and crosstalk. In this paper, we propose a novel time-multiplexed directional backlighting method to address the low-resolution limitations of II3D displays. In our proposed method, we first synchronize multi-angle dynamic collimated light with Elemental Image Array (EIA) refresh cycles. Then, we establish a pixel mapping model to precisely control the mapping process and effectively generate high-resolution 3D scenes. A multi-angle backlight II3D display prototype was built with our method. Additionally, we present a novel image quality evaluation method to assess the screen-door effect (SDE) present in the II3D. Experimental results demonstrate a ninefold increase in spatial resolution compared to single-angle directional backlighting methods, together with a significantly reduced SDE, which shows that our proposed method can effectively enhance both image resolution and image quality.

Digital holography imaging has been challenging due to the local maximum occurring through the auto-focusing process. It is also computationally extensive to search the auto focus region, costing large-range local searches to achieve the autofocus of the off-axis digital holography. With this research, a new efficient off-axis digital holography autofocus method was developed based on peak-support interval zoom searches shown able to balance computational efficiency and focusing accuracy. Specifically, an integrated high-sensitivity amplitude difference and searching interval-zoom strategy was adopted to rapidly narrow the focus searching range. It is followed by a second-order central difference amplitude difference method for further fine-tuning the optimal focal plane. Both simulation and experimental results demonstrate that compared with conventional autofocus methods, the proposed method achieves significant improvements in both the quality of the digital holography reconstruction and the efficiency of the autofocus of the off-axis digital holography.

In nanometric displacement measurement, laser interferometry exhibits high sensitivity to sub-period displacements. However, the interferometric signals are often incomplete in period and limited in amplitude, making them susceptible to high-frequency noise and drift. Therefore, an improved nanoscale displacement measurement method is proposed, combining band-limited dynamic-noise-assisted complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) with sinusoidal-regression demodulation. By constraining the noise band and adaptively tuning the noise amplitude during injection, the low-frequency trend component is effectively purified. Based on the prior parameters, such as laser wavelength and displacement rate, to constrain the fitting solution space, the linear least squares fitting method is used to process the trend signal to achieve high-precision phase recovery and quantitative displacement measurement. Experiments show that the method remains stable under electromagnetic interference and impulse noise conditions, improving the signal-to-noise ratio by about 2.6 dB. The measurement error is 1-3 nm within the 20-300 nm range, and the stability is improved by approximately 62%, while also satisfying real-time feasibility. Overall, the method achieves nm-level accuracy with enhanced robustness to industrial noise and real-time computational efficiency.

n advanced manufacturing and precision metrology, accurate measurement of the surface topography of each surface within transparent parallel optical components is crucial for controlling device performance. Multi-surface wavelength-tuned interferometry often suffers from spectral crowding and severe crosstalk due to the superposition of harmonic signals. Conventional FFT-based analysis is prone to spectral leakage under conditions of low signal-to-noise ratio (SNR) and non-integer period sampling, making it difficult to ensure measurement accuracy. To address these challenges, we propose a Hilbert-subspace Joint Parameter Estimation for Multi-surface Phase-shifting Interferometry (HJPE-MPI) algorithm. This algorithm utilizes the Hilbert transform to construct analytic signals, which are then used to form a Hankel matrix for subspace decomposition, enabling high-resolution frequency estimation of sampled interference signals. The estimated frequency offsets are subsequently used to correct the amplitude and phase of the harmonic components corresponding to each surface, thereby enabling joint and precise parameter localization. Simulation and experimental results demonstrate that the proposed method significantly outperforms conventional algorithms in the estimation accuracy of frequency, phase, and amplitude for multi-harmonic interference signals, and exhibits excellent repeatability and stability in measurements of a 30 mm-thick transparent plate.

Multi-surface wavelength-tuning interferometry is an essential technique for acquiring the nanoscale three-dimensional topography of optical parallel surfaces. Nevertheless, constrained by the superposition of multi-frequency signals, the finite sampling length, and the inherent discrete frequency resolution, the demodulation process is prone to spectral aliasing, leakage, and the picket-fence effect, which severely degrades the estimation accuracy of harmonic parameters. In this paper, we propose a complex-ratio correction method based on multi-snapshot spectral estimation algorithm (MSSE-mCRC). It utilizes a modified Singular Value Decomposition (SVD) to estimate the source number of superimposed interference harmonics, thereby assessing the sampling performance. By constructing a covariance matrix via multiple snapshots and employing eigendecomposition to separate the signal and noise subspaces, high-resolution spectral estimation is achieved. Subsequently, the Hanning-windowed Complex-Ratio Correction technique is applied to correct the amplitudes and phases. Simulation results demonstrate that the proposed algorithm achieves frequency and amplitude estimation errors on the order of 10-4 and 10-2, respectively, outperforming traditional FFT-based methods. Furthermore, in the repeated experiments performed on two different optical components, the maximum RMS, PVQ, and PV values of the measured surface profiles are 0.0111μm, 0.0454μm, and 0.0594μm for Experiment 1, and 0.0030μm, 0.0137μm, and 0.0308μm for Experiment 2, respectively. These results successfully validate the repeatability and stability of the proposed algorithm.

Fringe projection profilometry (FPP) is a wildly employed technique for three-dimension (3D) reconstruction. However, the inherent limited depth of field (DOF) of the FPP system extremely constrains its measurement capabilities, especially in scenes with large depth variations. Under out-of-focus conditions exceeding the system DOF, defocus blur in both the camera and projector introduces significant phase calculation errors, leading to degraded 3D reconstruction accuracy. To enhance the measurement capability and reconstruction accuracy of FPP system in complex scenes, overcoming the DOF limitation and compensating for defocus-induced errors are crucial. Thus, a parallel single-pixel imaging (PSI) based defocus compensation method is proposed in this paper. This approach constructs the light field based on the principle of PSI to describe the light transport process between the projector and camera pixels. From the light field, the point spread function (PSF) corresponding to different pixels in the image can be calculated for the FPP system, allowing deconvolution of the blurred image to compensate for phase solving error caused by defocusing. The experimental results show that the proposed method reduces the radius fitting error of the standard sphere in defocused state by up to approximately 17.3%, effectively compensating the phase solving error due to defocus. In the scene with the DOF that is about three times of the DOF of conventional scene, the proposed method preserves more details in the reconstruction results, thus improving the capability and robustness of 3D reconstruction.

Interferometry enables sub-nanometer accuracy in optical surface metrology. The wavelength-tuning phase-shifting interferometry facilitates simultaneous multi-surface measurements by leveraging the distinct frequencies induced by optical path differences. However, acquired interferograms contain superimposed interference patterns forming three harmonic signals, which cannot be processed by conventional signal separation methods due to single-channel constraints. Additionally, wavelength-tuning errors cause spectral leakage that degrades parameter reconstruction. To address issues presented, we propose a multi-surface wavelength-tuning interferometry algorithm using virtual interference channels for blind source separation and refined spectral analysis (MWI-VBSA). This method employs dual-criterion separation with orthogonalization to extract signal components, followed by spectral refinement for harmonic frequency and efficient phase demodulation functions. Comprehensive validation demonstrates significant improvements over FFT and multi-surface advanced iterative algorithm: simulations under varying SNR (25 similar to 50 dB) show reduced frequency errors and low surface reconstruction errors, while diverse sampling lengths/intervals confirm robust harmonic separation. Experimental measurements on 50 mm transparent plates achieve surface retrieval errors below 3.32 x 10(-3) lambda(0) for RMS, verifying consistent accuracy. Furthermore, the algorithm's robustness was decisively confirmed through a four-surface interferometry experiment, in which it successfully separated six superimposed harmonic signals. This outcome provides evidence of the algorithm's exceptional performance and reliability when processing complex signal superposition.

Accurate determination of the optimal reconstruction distance is essential for high-quality complex-amplitude reconstruction in off-axis digital holography. However, existing autofocus methods often struggle to balance focal-plane discriminability and search efficiency. To address this issue, we propose an autofocus method for off-axis digital holography based on the structure tensor eigenvalue anisotropy (STEA) and chaos-guided adaptive ripple search (CG-ARS). Specifically, the method constructs a focus evaluation function from the difference between the two eigenvalues of the Gaussian-regularized structure tensor, thereby improving the discrimination of the true focal plane by leveraging the increase in local structural directional anisotropy during focusing. In addition, we introduce CG-ARS to achieve efficient search and precise localization near the focal plane through the combination of aperiodic global exploration and adaptive step-size updating in the near-focus region. Numerical simulations and experimental results demonstrate that the proposed method is applicable to both amplitude and phase samples and can significantly reduce computational cost while maintaining focusing accuracy. The proposed method provides a robust and efficient solution for rapid autofocus in off-axis digital holography. It may also serve as a useful reference for numerical refocusing in other coherent computational imaging applications.

Murals face complex microcrack formation and deterioration due to diverse environmental conditions and time-dependent material behavior. Existing analytical methods remain limited in addressing these challenges. This paper proposes a cross-scale analysis approach that integrates digital holographic non-destructive testing with molecular dynamics simulation. An in-situ digital holographic detection system was developed to achieve sub-micron three-dimensional characterization of microcracks on simulated mural samples. Molecular dynamics models were constructed based on extracted feature parameters to dynamically simulate microcrack propagation. Aging experiments were conducted to validate the simulation results, enabling cross-scale correlation and mutual verification between experimental observations and mechanistic simulations. The study demonstrates that this method successfully establishes a qualitative link between microscopic morphological damage in murals and molecular-scale evolution. It offers a new technical pathway for tracing deterioration mechanisms, predicting evolution trends, and supporting conservation assessments of cultural heritage.

To realize four-surface wavelength-tuning phase-shifting interferometry under blind cavity length and sampling frequency, an adaptive wavelength-tuning phase-shifting interferometry algorithm is proposed to obtain the front surface profile, rear surface profile, thickness variation, and refractive index inhomogeneity of the transparent parallel plate. This paper iteratively cycles through cavity lengths and sampling frequencies, quantifying the phase-shifting algorithm's solving error under each parameter combination. By introducing an error threshold to evaluate solving performance, it identifies all valid combinations of cavity lengths and sampling frequencies. Numerical simulations were used to calculate the combinations of cavity length and sampling frequency, and the validity was verified. Repetitive experiments were conducted under different combinations of blind cavity length and sampling frequency on a transparent plate with a thickness of 30 mm. The results show that the self-assessment index for interferometry error is on the order of 0.0034 lambda 0. The maximum deviations of the average values of PVQ for the reconstructed front surface profile, rear surface profile, thickness variation, and refractive index inhomogeneity of the transparent parallel plate are 0.001 lambda 0, 0.0048 lambda 0, 0.0041 lambda 0, and 1.3717 x 10- 7, respectively, proving the effectiveness of the proposed algorithm.

The detection and degradation analysis of subsurface microcracks in mural paintings remain challenging due to their inhomogeneous multilayered structure and complex deterioration mechanisms. In this study, we propose a multimodal stepwise method for three-dimensional characterization and cross-scale degradation analysis by integrating digital holography (DH), infrared thermography (IRT), acoustic excitation (AE), and molecular dynamics (MD) simulations. In the first step, an adjustable field-of-view (FOV) digital holographic system is developed to capture subsurface deformation under acoustic excitation, enabling high-resolution planar characterization of subsurface microcracks. Infrared thermography is then employed to estimate crack depth through an inverse thermal model, achieving full three-dimensional reconstruction of crack geometry. Based on the reconstructed structures, MD simulations are conducted to investigate the evolution of stress, bond breaking, and crack propagation under varying temperature and humidity conditions, with particular emphasis on water molecule migration and chemically induced degradation. The results demonstrate that environmental factors promote stress concentration and material embrittlement at crack tips, leading to secondary microcrack formation and progressive deterioration. Experimental aging tests show strong agreement with simulation results, validating the proposed methodology. This work establishes a unified “characterization–simulation–validation” paradigm, providing new insights into the mechanisms of mural degradation and offering a robust framework for non-destructive evaluation and preventive conservation of multilayer cultural heritage materials.

High-scattering ceramics, such as thermal barrier coatings (TBCs), are primary targets for mid-infrared optical coherence tomography (MIR-OCT). Compared with conventional near-infrared OCT, MIR-OCT systems leverage long-wavelength light sources to suppress optical scattering significantly, achieving substantial breakthroughs in both penetration depth and signal-to-noise ratio (SNR). However, current optimizations of the system’s anti-scattering performance primarily focus on a single approach: increasing the wavelength to reduce scattering. This study demonstrates that, in addition to the inherent advantages of longer wavelengths, the anti-scattering performance of MIR‑OCT can be further enhanced by discriminating the spatial and polarization characteristics of backscattered light. We first determined the scattering and absorption coefficients of TBCs in the mid-infrared range using a four-flux Kubelka-Munk (4F-KM) method. These parameters were integrated into the polarimetric Monte Carlo multi-layer (PMCML) model to study the relationship between the scattering frequency of backscattered photons and their spatial distribution and polarization evolution. Based on this analysis, two anti-scattering strategies are proposed: spatial filtering and polarization-based filtering. Simulation results demonstrate that both approaches effectively suppress multiply scattered photons. Experiments further confirm that spatial filtering, as one of the two strategies, substantially reduces scattering-induced background noise, validating its effectiveness in improving MIR‑OCT system performance.

This study addresses the issue of image blurring caused by high-speed motion in industrial chip flying shot scenarios by proposing a collaborative optimization framework Adaptive Deconvolution Wiener (ADWiener) based on a multi-scale attention kernel estimation network (PC-MANet) and an improved adaptive frequency Wiener filter (AFWF). The PC-MANet achieves precise multi-scale blur kernel estimation through lightweight dilated convolution modules and a dual-attention mechanism combining gated fusion of Mixed Local Channel Attention (MLCA) and (Multi-Head Self-Attention) MHSA. The AFWF innovatively introduces dynamic regularization and phase correction techniques, significantly enhancing texture fidelity while reducing ringing artifacts. Experiments on a chip motion blur dataset demonstrate that ADWiener achieves a PSNR of 23.883 dB, an SSIM of 0.716, and real-time performance of 34 fps on the test set, exhibiting strong robustness in multi-directional and high-speed chip motion scenarios. This provides a high-precision, low-latency solution for industrial visual inspection.

Digital holographic microscopy (DHM) enables non-contact, label-free, high-resolution quantitative phase imaging. However, unavoidable wavefront aberrations in the optical system introduce severe artifacts and distortions into the reconstructed phase, thereby limiting the accuracy of quantitative measurements. To address this issue, we propose a phase aberration compensation method for DHM based on background adaptive segmentation and iterative constrained fitting. First, a least-squares (LS) method is employed to coarsely remove the dominant low-order aberrations. Subsequently, an initial background mask is constructed from the magnitude of the phase gradient. Background candidate regions with good connectivity are then obtained by combining Otsu’s adaptive thresholding with morphological operations. These regions are then embedded into a level-set framework to achieve adaptive and refined segmentation of the object and background. At each iteration, only background pixels are used to solve a constrained LS fitting problem to estimate the residual aberrations. The standard deviation (STD) of the background phase is used as the convergence criterion, thereby enabling adaptive compensation of phase aberrations. Simulation and experimental results demonstrate that the proposed method significantly outperforms conventional approaches in terms of phase aberration compensation accuracy and robustness. It effectively recovers and corrects system aberrations for various types of samples, yielding high-quality quantitative phase images.

By integrating step-size decreasing search with particle swarm optimization algorithms, we resolved focusing errors in digital holographic reconstruction, achieving automatic focusing and clear reconstruction of holographic images.

Efficient and precise measurements of optics with parallel surfaces are crucial for ensuring component performance. However, current interferometry faces challenges from an uncertain number of interference sources and adverse effects on parameter demodulation due to sampling errors and noise. To address these issues, we first use the minimum description length (MDL) criterion on a Hankel matrix of intensity signals to estimate the number of effective surfaces by minimizing an information-theoretic cost function. Then, variational mode decomposition (VMD) iteratively extracts the superimposed signal into interference modal components (IMCs) with distinct frequencies. Frequency estimation employs a zero-padded Fourier transform with a Blackman window, followed by phase and surface reconstruction. In error analysis, under signal-to-noise ratio (SNR) = 30-50 dB and up to 20% phase-shifting errors, frequency solution errors remain below 0.7%. The Monte Carlo trials show low mean and standard deviation (STD) in solution errors. Repeated measurements on a transparent plate yield stable profiles with minimal peak-to-valley (PV), root mean square (RMS), and PVQ errors.

In this work, we propose a high-precision off-axis holographic reconstruction method that integrates Kronecker interpolation with the fractional Fourier transform (FRFT), which can flexibly rotate signals between the spatial and frequency domains, enabling high-quality reconstruction in off-axis digital holography. First, Kronecker interpolation is employed to enhance the spatial sampling density of the hologram and to strengthen its spectral characteristics, thereby effectively suppressing spectral aliasing induced by the zero-order diffraction. Subsequently, taking advantage of the rotational additivity property of the FRFT, a Grey Wolf Optimizer (GWO) is introduced to search for the optimal fractional order adaptively. Finally, amplitude and phase distributions are reconstructed with high fidelity using the optimized FRFT. Both simulation and experimental results demonstrate that the proposed method substantially improves hologram reconstruction quality. The approach provides an effective and scalable solution for enhancing off-axis digital holographic imaging, with significant implications for precise micro/nano-scale metrology and high-resolution biological imaging.

Shape reconstruction and force measurement of surgical diagnostic tools are crucial to ensure surgical safety. This study focuses on the strain transfer of an overall flexible fiber Bragg grating (FBG) shape sensor. A theoretical equation of strain transfer was established, with the fiber and bonding layers as the core structures. The response characteristics of the fibers under axial and radial forces were investigated. Through finite element simulations, the strain values of the bonding and fiber layers were analyzed and compared with the theoretical values to determine the influence of the bonding layer thickness, elastic modulus, encapsulation length, and Poisson’s ratio on the strain transfer. A 260-mm-long, 2-mm-diameter, substrate-free FBG shape sensor was encapsulated, and axial/radial force sensing experiments were conducted to verify the theoretical force-wavelength change model. Additionally, shape reconstruction experiments were performed on a flat surface, achieving a 3.68

Digital acousto-optic holography (DAOH) integrates ultrasonic excitation with digital holographic measurement to enable full-field, non-contact detection of subsurface defects in opaque solids. This review outlines the operating principles of DAOH and compares its two primary optical recording modes-television holography for displacement measurement and digital shearography for strain-gradient detection. The complex acoustic field obtained at the surface is numerically propagated into the material interior using Rayleigh-Sommerfeld diffraction, enabling volumetric reconstruction of internal discontinuities. Detection performance, including fringe visibility, sensitivity, robustness, and depth resolution, is systematically evaluated. Current challenges related to mode conversion, attenuation, and material heterogeneity are discussed, along with recent advances in high-speed acquisition, multi-wavelength deformation measurement, and dual-mode optical architectures. Prospects such as multi-mode excitation, super-resolution acousto-optic imaging, and physics-informed reconstruction are highlighted as key pathways toward real-time, high-fidelity three-dimensional non-destructive testing imaging.

The coupling alignment of single-mode fibers demands stringent optical mode matching accuracy, where micro-displacements and angular tilts drastically degrade coupling efficiency. Furthermore, traditional alignment processes frequently suffer from insufficient precision, slow convergence, and poor algorithmic stability. To overcome these limitations, this paper proposes a sensitivity matrix-based geometric particle swarm optimization (SG-PSO) algorithm. By introducing a sensitivity matrix to explicitly map the physical relationship between mechanical degrees of freedom and wavefront distortions, the proposed method provides clear geometric guidance to the particle swarm optimizer. The sensitivity matrix is seamlessly embedded into the PSO update rule to provide a deterministic search direction. The alignment process is comprehensively simulated using Zemax optical software and validated through physical experiments employing a 618 nm laser. Experimental results demonstrate that SG-PSO achieves a 50% reduction in eccentricity and tilt errors compared to traditional methods. Notably, the SG-PSO algorithm rapidly approaches the theoretical alignment limit, converging to a highly accurate solution within 15 iterations. Ultimately, SG-PSO exhibits substantial advantages in both positioning precision and convergence speed, offering a robust and highly efficient optimization paradigm for multi-degree-of-freedom fiber coupling alignment.