Optics and Lasers in Engineering最新文献_第7页

DeepBessel: deep learning-based full-field vibration profilometry using single-shot time-averaged interference microscopy DeepBessel：使用单次时间平均干涉显微镜的基于深度学习的全场振动轮廓术

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-31 DOI: 10.1016/j.optlaseng.2025.109588

Maria Cywińska, Wiktor Forjasz, Emilia Wdowiak, Michał Józwik, Adam Styk, Krzysztof Patorski, Maciej Trusiak

Full-field vibration profilometry is essential for dynamic characterizing microelectromechanical systems (MEMS/MOEMS). Time-averaged interferometry (TAI) encodes spatial information about MEMS/MOEMS vibration amplitude in the interferogram’s amplitude modulation using Bessel function (besselogram). Classical approaches for interferogram analysis are specialized for cosine function fringe patterns and therefore introduce reconstruction errors for besselogram decoding. This paper presents the DeepBessel: a deep learning-based approach for single-shot TAI interferogram analysis. A convolutional neural network (CNN) was trained using synthetic data, where the input consisted of besselograms, and the output corresponded to the underlying vibration amplitude distribution. Numerical validation and experimental testing demonstrated that DeepBessel significantly reduces reconstruction errors compared to the state-of-the-art approaches, e.g., Hilbert Spiral Transform (HST) method. The proposed network effectively mitigates errors caused by the mismatch between the Bessel and cosine functions. The results indicate that deep learning can improve the accuracy of full-field vibration measurements, offering new possibilities for optical metrology in MEMS/MOEMS applications.

全场振动轮廓测量法是动态表征微机电系统（MEMS/MOEMS）的关键。时间平均干涉法（TAI）利用贝塞尔函数（Bessel function, besselogram）对干涉图的振幅调制中MEMS/MOEMS振动振幅的空间信息进行编码。干涉图分析的经典方法专门用于余弦函数条纹图案，因此引入了贝塞图解码的重建误差。提出了一种基于深度学习的单次TAI干涉图分析方法——DeepBessel。使用合成数据训练卷积神经网络（CNN），其中输入由besselogram组成，输出对应底层振动幅值分布。数值验证和实验测试表明，与希尔伯特螺旋变换（HST）方法等最先进的方法相比，DeepBessel显著降低了重建误差。该网络有效地减轻了贝塞尔函数和余弦函数不匹配所引起的误差。结果表明，深度学习可以提高全场振动测量的精度，为MEMS/MOEMS应用中的光学测量提供了新的可能性。

{"title":"DeepBessel: deep learning-based full-field vibration profilometry using single-shot time-averaged interference microscopy","authors":"Maria Cywińska, Wiktor Forjasz, Emilia Wdowiak, Michał Józwik, Adam Styk, Krzysztof Patorski, Maciej Trusiak","doi":"10.1016/j.optlaseng.2025.109588","DOIUrl":"10.1016/j.optlaseng.2025.109588","url":null,"abstract":"<div><div>Full-field vibration profilometry is essential for dynamic characterizing microelectromechanical systems (MEMS/MOEMS). Time-averaged interferometry (TAI) encodes spatial information about MEMS/MOEMS vibration amplitude in the interferogram’s amplitude modulation using Bessel function (besselogram). Classical approaches for interferogram analysis are specialized for cosine function fringe patterns and therefore introduce reconstruction errors for besselogram decoding. This paper presents the DeepBessel: a deep learning-based approach for single-shot TAI interferogram analysis. A convolutional neural network (CNN) was trained using synthetic data, where the input consisted of besselograms, and the output corresponded to the underlying vibration amplitude distribution. Numerical validation and experimental testing demonstrated that DeepBessel significantly reduces reconstruction errors compared to the state-of-the-art approaches, e.g., Hilbert Spiral Transform (HST) method. The proposed network effectively mitigates errors caused by the mismatch between the Bessel and cosine functions. The results indicate that deep learning can improve the accuracy of full-field vibration measurements, offering new possibilities for optical metrology in MEMS/MOEMS applications.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109588"},"PeriodicalIF":3.7,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885328","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Hyperspectral images 3D reconstruction based on structure-from-motion and multi-view stereo 基于运动结构和多视点立体的高光谱图像三维重建

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-31 DOI: 10.1016/j.optlaseng.2025.109578

Chao Liu , Minghui Gu , Huansong Huang , Xin Zhang , Jiguo Li , Qingbin Jiao , Liang Xu , Mingyu Yang , Xin Tan

Hyperspectral 3D imaging technology aims to obtain 3D models with hyperspectral reflectance textures, holds significant potential in fields such as medical imaging, precision agriculture, and optical sorting. However, the generation and application of hyperspectral 3D models face numerous challenges due to the complexities of light propagation in intricate environments, particularly in interactions with various media. In order to achieve a one-to-one mapping between hyperspectral data and point cloud coordinates, many researchers have developed different hyperspectral three-dimensional imaging systems to address the problem. In this paper, we propose a framework based on the Structure-from-Motion (SFM) and Multi-View Stereo (MVS) to process the multi-view hyperspectral images collected by an internal push-broom hyperspectral camera. In the framework, an image segmentation module is introduced to suppress the background noise during the target reconstruction process. In the Structure-from-Motion algorithm, a matching strategy for circular scanning mode is adopted to effectively alleviate the influence of repetitive textures and large-span mismatches on camera pose estimation, thereby improving the pointcloud stitching accuracy. And we utilize the multi-channel redundant information to enhance the robustness of stereo matching depth estimation. Additionally, we introduce a fusion mechanism that incorporates mask-based noise suppression and statistical reflectance integration to enhance geometric accuracy and spectral fidelity. The hyperspectral reflectance is not severely distorted through aggregating the spectrum of redundant observed points. Experimental results on real-world data collected by our system demonstrate that the effectiveness of our method to handle spectral variability and geometric complexity.

高光谱3D成像技术旨在获得具有高光谱反射纹理的3D模型，在医学成像、精准农业和光学分选等领域具有重要潜力。然而，由于光在复杂环境中传播的复杂性，特别是与各种介质的相互作用，高光谱3D模型的生成和应用面临许多挑战。为了实现高光谱数据与点云坐标的一对一映射，许多研究人员开发了不同的高光谱三维成像系统来解决这一问题。本文提出了一种基于运动结构（SFM）和多视点立体（MVS）的框架来处理内部推扫帚式高光谱相机采集的多视点高光谱图像。在该框架中，引入了图像分割模块来抑制目标重构过程中的背景噪声。在Structure-from-Motion算法中，采用圆形扫描模式的匹配策略，有效缓解了重复纹理和大跨度错配对相机姿态估计的影响，从而提高了点云拼接精度。利用多通道冗余信息增强立体匹配深度估计的鲁棒性。此外，我们还引入了一种融合了基于掩模的噪声抑制和统计反射率集成的融合机制，以提高几何精度和光谱保真度。通过对冗余观测点的光谱进行聚合，使高光谱反射率没有严重失真。实验结果表明，该方法能够有效地处理光谱变异性和几何复杂性。

{"title":"Hyperspectral images 3D reconstruction based on structure-from-motion and multi-view stereo","authors":"Chao Liu , Minghui Gu , Huansong Huang , Xin Zhang , Jiguo Li , Qingbin Jiao , Liang Xu , Mingyu Yang , Xin Tan","doi":"10.1016/j.optlaseng.2025.109578","DOIUrl":"10.1016/j.optlaseng.2025.109578","url":null,"abstract":"<div><div>Hyperspectral 3D imaging technology aims to obtain 3D models with hyperspectral reflectance textures, holds significant potential in fields such as medical imaging, precision agriculture, and optical sorting. However, the generation and application of hyperspectral 3D models face numerous challenges due to the complexities of light propagation in intricate environments, particularly in interactions with various media. In order to achieve a one-to-one mapping between hyperspectral data and point cloud coordinates, many researchers have developed different hyperspectral three-dimensional imaging systems to address the problem. In this paper, we propose a framework based on the Structure-from-Motion (SFM) and Multi-View Stereo (MVS) to process the multi-view hyperspectral images collected by an internal push-broom hyperspectral camera. In the framework, an image segmentation module is introduced to suppress the background noise during the target reconstruction process. In the Structure-from-Motion algorithm, a matching strategy for circular scanning mode is adopted to effectively alleviate the influence of repetitive textures and large-span mismatches on camera pose estimation, thereby improving the pointcloud stitching accuracy. And we utilize the multi-channel redundant information to enhance the robustness of stereo matching depth estimation. Additionally, we introduce a fusion mechanism that incorporates mask-based noise suppression and statistical reflectance integration to enhance geometric accuracy and spectral fidelity. The hyperspectral reflectance is not severely distorted through aggregating the spectrum of redundant observed points. Experimental results on real-world data collected by our system demonstrate that the effectiveness of our method to handle spectral variability and geometric complexity.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109578"},"PeriodicalIF":3.7,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885323","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A novel method of designing integrating mirrors for two-dimensional customized lasers in surface hardening process 一种用于二维定制激光器表面硬化的积分镜设计新方法

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-30 DOI: 10.1016/j.optlaseng.2025.109580

Shuo Li , Yu Long , Lirong Zhong , Yajun Zhang , Mingfeng Liu , Yihu Zhang , Qi Gao , Xiahui Tang , Yu Xiao , Yingxiong Qin

Two-dimensional customized lasers obtained by conventional integrating mirrors are widely employed in industrial applications such as hardening, cladding, and surface heating, because of their high energy output efficiency and large radiation area. However, the existing two-dimensional customized lasers exhibit limited spot size along the length direction and an energy distribution confinement of the Gaussian mode along the width direction. In this work, we propose a novel generatrix design method, which is composed of alternating convex and concave high-order curve segments, to enhance energy uniformity and spot size along the length direction. Besides, a translation scheme applicable to all generatrixes is proposed to adjust the energy distribution along the width direction. Additionally, improved surface smoothness enhances manufacturability, allowing higher input power tolerance with minimal output loss. Here, we presented a 100.40-mm-long stripe-shaped laser, as well as a 21.17-mm-long dual-stripe-shaped laser with a peak interval of 3.98 mm. The energy distribution simulation results demonstrate the effectiveness of the novel beam shaping method and highlights its strong potential for high-power laser processing and other advanced laser applications.

传统积分镜获得的二维定制激光器因其能量输出效率高、辐射面积大而广泛应用于硬化、熔覆、表面加热等工业领域。然而，现有的二维定制激光器在长度方向上光斑尺寸有限，在宽度方向上能量分布受高斯模式的限制。在这项工作中，我们提出了一种新的母线设计方法，该方法由凹凸交替的高阶曲线段组成，以提高沿长度方向的能量均匀性和光斑尺寸。此外，提出了一种适用于所有母线的平移方案，以调整沿宽度方向的能量分布。此外，改进的表面光洁度提高了可制造性，以最小的输出损耗实现更高的输入功率公差。在这里，我们提出了一个长100.40 mm的条形激光器，以及一个长21.17 mm的双条形激光器，峰值间隔为3.98 mm。能量分布仿真结果证明了这种新型光束整形方法的有效性，并突出了其在高功率激光加工和其他先进激光应用中的强大潜力。

{"title":"A novel method of designing integrating mirrors for two-dimensional customized lasers in surface hardening process","authors":"Shuo Li , Yu Long , Lirong Zhong , Yajun Zhang , Mingfeng Liu , Yihu Zhang , Qi Gao , Xiahui Tang , Yu Xiao , Yingxiong Qin","doi":"10.1016/j.optlaseng.2025.109580","DOIUrl":"10.1016/j.optlaseng.2025.109580","url":null,"abstract":"<div><div>Two-dimensional customized lasers obtained by conventional integrating mirrors are widely employed in industrial applications such as hardening, cladding, and surface heating, because of their high energy output efficiency and large radiation area. However, the existing two-dimensional customized lasers exhibit limited spot size along the length direction and an energy distribution confinement of the Gaussian mode along the width direction. In this work, we propose a novel generatrix design method, which is composed of alternating convex and concave high-order curve segments, to enhance energy uniformity and spot size along the length direction. Besides, a translation scheme applicable to all generatrixes is proposed to adjust the energy distribution along the width direction. Additionally, improved surface smoothness enhances manufacturability, allowing higher input power tolerance with minimal output loss. Here, we presented a 100.40-mm-long stripe-shaped laser, as well as a 21.17-mm-long dual-stripe-shaped laser with a peak interval of 3.98 mm. The energy distribution simulation results demonstrate the effectiveness of the novel beam shaping method and highlights its strong potential for high-power laser processing and other advanced laser applications.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109580"},"PeriodicalIF":3.7,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885325","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A comparative review and benchmark for deep learning based digital image correlation method 基于深度学习的数字图像相关方法的比较综述和基准

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-30 DOI: 10.1016/j.optlaseng.2025.109592

Lianpo Wang , Han Wu

After more than forty years of development, the theory and application of DIC methods have been extensively researched and developed. But in the past five years, the introduction of deep learning has brought new implementation solutions and development opportunities to the field of DIC. The deep learning based DIC (DL-DIC) method not only has higher computational efficiency and easier operation, but also can measure more complex high-frequency deformations. In order to help DIC researchers quickly get started with DL-DIC methods, this paper provides a comparative review of DL-DIC for the first time. This review not only comprehensively reviews the network structure, loss function, and dataset of existing DL-DIC methods, but also divides the existing network from the perspective of measurable deformation range, constructs datasets with different deformation ranges, and conducts comparative experiments on existing DL-DIC methods. Based on comparative experiments, we provide recommendations for selecting DL-DIC networks, avoiding repeated attempts by DIC researchers and users. We also open sourced the dataset and all the DL-DIC network code of this review to promote the development of the DL-DIC field.

经过四十多年的发展，DIC方法的理论和应用得到了广泛的研究和发展。但在过去的五年中，深度学习的引入为DIC领域带来了新的实现方案和发展机会。基于深度学习的DIC （DL-DIC）方法不仅具有更高的计算效率和更简单的操作，而且可以测量更复杂的高频变形。为了帮助DIC研究者快速上手DL-DIC方法，本文首次对DL-DIC方法进行了比较综述。本文不仅对现有DL-DIC方法的网络结构、损失函数和数据集进行了全面的综述，而且从可测量变形范围的角度对现有网络进行了划分，构建了不同变形范围的数据集，并对现有DL-DIC方法进行了对比实验。在对比实验的基础上，我们提出了选择DL-DIC网络的建议，避免了DIC研究人员和用户的重复尝试。我们还开放了本综述的数据集和所有DL-DIC网络代码，以促进DL-DIC领域的发展。

{"title":"A comparative review and benchmark for deep learning based digital image correlation method","authors":"Lianpo Wang , Han Wu","doi":"10.1016/j.optlaseng.2025.109592","DOIUrl":"10.1016/j.optlaseng.2025.109592","url":null,"abstract":"<div><div>After more than forty years of development, the theory and application of DIC methods have been extensively researched and developed. But in the past five years, the introduction of deep learning has brought new implementation solutions and development opportunities to the field of DIC. The deep learning based DIC (DL-DIC) method not only has higher computational efficiency and easier operation, but also can measure more complex high-frequency deformations. In order to help DIC researchers quickly get started with DL-DIC methods, this paper provides a comparative review of DL-DIC for the first time. This review not only comprehensively reviews the network structure, loss function, and dataset of existing DL-DIC methods, but also divides the existing network from the perspective of measurable deformation range, constructs datasets with different deformation ranges, and conducts comparative experiments on existing DL-DIC methods. Based on comparative experiments, we provide recommendations for selecting DL-DIC networks, avoiding repeated attempts by DIC researchers and users. We also open sourced the dataset and all the DL-DIC network code of this review to promote the development of the DL-DIC field.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109592"},"PeriodicalIF":3.7,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885322","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Dynamic calibration and reconstruction for two-dimensional spatial modulation snapshot imaging polarimeter under misalignment errors: Principle, Simulation, and experiment 二维空间调制快照成像偏振计在不对准误差下的动态校准与重建：原理、仿真与实验

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-30 DOI: 10.1016/j.optlaseng.2025.109589

An Pan , Yilin Zhang , Chenling Jia , Qizhi Cao , Jing Zhang

The Two-Dimensional spatial modulation snapshot imaging polarimeter (SMSIP) can obtain the full Stokes components of a target in a single shot. However, dynamic errors in modulation phase and angular misalignment errors in core optical components can lead to carrier frequency shift, spurious peaks, and channel crosstalk, significantly degrading the accuracy of the Stokes-vector reconstruction. This study modeled the mechanism of misalignment errors and quantitatively characterized their effect on the Fourier-domain channel data. A dynamic calibration and error-correction method for polarization reconstruction under misalignment errors is proposed, which allows simultaneous target measurement and phase calibration while compensating for all angular alignment-induced inaccuracies. Both simulations and experimental results demonstrate that the proposed method improves the reconstruction accuracy of the full Stokes vector while maintaining single-exposure efficiency, with a significant enhancement in the reconstruction of the S₂ component. Furthermore, within a certain range, the corrective performance improves markedly with increasing signal-to-noise ratio. This research significantly advances the engineering process of SMSIP systems.

二维空间调制快照成像偏振计（SMSIP）可以在一次拍摄中获得目标的全部斯托克斯分量。然而，调制相位的动态误差和核心光学元件的角度失调误差会导致载波频移、杂散峰和信道串扰，从而显著降低stokes矢量重建的精度。本研究模拟了不对准误差的机制，并定量表征了它们对傅里叶域信道数据的影响。提出了一种用于偏振重建的动态校准和误差校正方法，该方法可以同时进行目标测量和相位校准，同时补偿所有角对准引起的误差。仿真和实验结果表明，该方法在保持单次曝光效率的同时，提高了全Stokes矢量的重建精度，其中S₂分量的重建效果显著增强。此外，在一定范围内，校正性能随信噪比的增加而显著提高。该研究对SMSIP系统的工程化进程具有重要的推动作用。

{"title":"Dynamic calibration and reconstruction for two-dimensional spatial modulation snapshot imaging polarimeter under misalignment errors: Principle, Simulation, and experiment","authors":"An Pan , Yilin Zhang , Chenling Jia , Qizhi Cao , Jing Zhang","doi":"10.1016/j.optlaseng.2025.109589","DOIUrl":"10.1016/j.optlaseng.2025.109589","url":null,"abstract":"<div><div>The Two-Dimensional spatial modulation snapshot imaging polarimeter (SMSIP) can obtain the full Stokes components of a target in a single shot. However, dynamic errors in modulation phase and angular misalignment errors in core optical components can lead to carrier frequency shift, spurious peaks, and channel crosstalk, significantly degrading the accuracy of the Stokes-vector reconstruction. This study modeled the mechanism of misalignment errors and quantitatively characterized their effect on the Fourier-domain channel data. A dynamic calibration and error-correction method for polarization reconstruction under misalignment errors is proposed, which allows simultaneous target measurement and phase calibration while compensating for all angular alignment-induced inaccuracies. Both simulations and experimental results demonstrate that the proposed method improves the reconstruction accuracy of the full Stokes vector while maintaining single-exposure efficiency, with a significant enhancement in the reconstruction of the S₂ component. Furthermore, within a certain range, the corrective performance improves markedly with increasing signal-to-noise ratio. This research significantly advances the engineering process of SMSIP systems.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109589"},"PeriodicalIF":3.7,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885324","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Hyperspectral image spectral-spatial classification with Gaussian processes and Markov random field 基于高斯过程和马尔可夫随机场的高光谱图像光谱空间分类

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-29 DOI: 10.1016/j.optlaseng.2025.109568

Yaqiu Zhang, Quanhua Zhao, Yu Li, Xueliang Gong

This paper proposes a new spectral-spatial approach for hyperspectral image (HSI) classification based on sparse inducing points variational Gaussian processes (SVGP) and spatially adaptive regularization Markov random field (SpAR), named SVGP-SpAR. Firstly, the given image is modeled as a spectral random field with band correlation using SVGP. The posterior probability is approximated through sparse method and variational inference, and the corresponding parameters of the SVGP are obtained using the adaptive sign momentum (ASM) unconstrained stochastic optimizer. Secondly, due to the uncertainty of the spatial distribution properties of the object, the point estimates of the latent function predicted by SVGP contain significant random additive noise. Consequently, this paper introduces the SpAR model to construct the spatial random field, applying spatial priors to the point estimates. Under the Bayesian framework, SpAR transforms the probabilistic model into a constrained regularization model to complete the SVGP-SpAR classification. Experimental results on three different HSI datasets demonstrate that the proposed SVGP-SpAR method outperforms existing methods in terms of classification accuracy across both spectral and spatial dimensions. Additionally, the model and parameters exhibit clear physical significance, making them easier to understand and interpret.

提出了一种基于稀疏诱导点变分高斯过程（SVGP）和空间自适应正则化马尔可夫随机场（SpAR）的高光谱图像空间分类新方法，命名为SVGP-SpAR。首先，利用SVGP将给定图像建模为带相关的光谱随机场；通过稀疏方法和变分推理逼近后验概率，并利用自适应符号动量（ASM）无约束随机优化器获得相应的SVGP参数。其次，由于目标空间分布特性的不确定性，svm预测的隐函数点估计含有显著的随机加性噪声。因此，本文引入SpAR模型构建空间随机场，将空间先验应用于点估计。SpAR在贝叶斯框架下，将概率模型转化为约束正则化模型，完成SVGP-SpAR分类。在三个不同的HSI数据集上的实验结果表明，所提出的SVGP-SpAR方法在光谱和空间维度上的分类精度都优于现有方法。此外，模型和参数表现出明确的物理意义，使它们更容易理解和解释。

{"title":"Hyperspectral image spectral-spatial classification with Gaussian processes and Markov random field","authors":"Yaqiu Zhang, Quanhua Zhao, Yu Li, Xueliang Gong","doi":"10.1016/j.optlaseng.2025.109568","DOIUrl":"10.1016/j.optlaseng.2025.109568","url":null,"abstract":"<div><div>This paper proposes a new spectral-spatial approach for hyperspectral image (HSI) classification based on sparse inducing points variational Gaussian processes (SVGP) and spatially adaptive regularization Markov random field (SpAR), named SVGP-SpAR. Firstly, the given image is modeled as a spectral random field with band correlation using SVGP. The posterior probability is approximated through sparse method and variational inference, and the corresponding parameters of the SVGP are obtained using the adaptive sign momentum (ASM) unconstrained stochastic optimizer. Secondly, due to the uncertainty of the spatial distribution properties of the object, the point estimates of the latent function predicted by SVGP contain significant random additive noise. Consequently, this paper introduces the SpAR model to construct the spatial random field, applying spatial priors to the point estimates. Under the Bayesian framework, SpAR transforms the probabilistic model into a constrained regularization model to complete the SVGP-SpAR classification. Experimental results on three different HSI datasets demonstrate that the proposed SVGP-SpAR method outperforms existing methods in terms of classification accuracy across both spectral and spatial dimensions. Additionally, the model and parameters exhibit clear physical significance, making them easier to understand and interpret.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109568"},"PeriodicalIF":3.7,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885321","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A study on the polarization aberration correction method of the micro space debris calibration device 微空间碎片定标装置偏振像差校正方法研究

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-29 DOI: 10.1016/j.optlaseng.2025.109585

Linghao Wu , Yu Zhang , Zimo Yang

The polarization aberration in the Microscopic Defect Detector (MCD) causes the stray light polarization field and the calibration field to intertwine, making them difficult to distinguish and thereby limiting the improvement of calibration accuracy. To address this issue, this study first analyzes the generation mechanism of polarization aberration and then investigates how the optical structural parameters of the MCD influence it. A combined polarization balance optimization method for Polarizing Beam Splitter (PBS) films is proposed to mitigate the polarization aberration induced by structural defects in the PBS films, thereby suppressing the stray light polarization field. Simulation results demonstrate that after applying the proposed optimization, the two-term attenuation and phase delay are reduced by a factor of 1.65 and 1.26, respectively. Experimental validation shows that the calibration error of the MCD is decreased by a factor of 1.27 when the optimized PBS films are employed. This research holds significant importance for the precise calibration of target fields in high-end aerospace and spacecraft attitude and orbit measurement systems.

显微缺陷检测器（MCD）中存在的偏振像差导致杂散光偏振场与标定场交织在一起，难以区分，从而限制了标定精度的提高。为了解决这一问题，本研究首先分析了偏振像差的产生机理，然后研究了MCD光学结构参数对偏振像差的影响。针对极化分束器（PBS）薄膜结构缺陷引起的偏振像差，提出了一种组合式偏振平衡优化方法，从而抑制杂散光偏振场。仿真结果表明，采用该优化方法后，两项衰减和相位延迟分别降低了1.65和1.26倍。实验验证表明，使用优化后的PBS膜，MCD的校准误差降低了1.27倍。该研究对高端航天航天器姿态轨道测量系统中目标场的精确标定具有重要意义。

{"title":"A study on the polarization aberration correction method of the micro space debris calibration device","authors":"Linghao Wu , Yu Zhang , Zimo Yang","doi":"10.1016/j.optlaseng.2025.109585","DOIUrl":"10.1016/j.optlaseng.2025.109585","url":null,"abstract":"<div><div>The polarization aberration in the Microscopic Defect Detector (MCD) causes the stray light polarization field and the calibration field to intertwine, making them difficult to distinguish and thereby limiting the improvement of calibration accuracy. To address this issue, this study first analyzes the generation mechanism of polarization aberration and then investigates how the optical structural parameters of the MCD influence it. A combined polarization balance optimization method for Polarizing Beam Splitter (PBS) films is proposed to mitigate the polarization aberration induced by structural defects in the PBS films, thereby suppressing the stray light polarization field. Simulation results demonstrate that after applying the proposed optimization, the two-term attenuation and phase delay are reduced by a factor of 1.65 and 1.26, respectively. Experimental validation shows that the calibration error of the MCD is decreased by a factor of 1.27 when the optimized PBS films are employed. This research holds significant importance for the precise calibration of target fields in high-end aerospace and spacecraft attitude and orbit measurement systems.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109585"},"PeriodicalIF":3.7,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multi-frame sequence-based multi-scale registration for dynamic fluid images 基于多帧序列的动态流体图像多尺度配准

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-29 DOI: 10.1016/j.optlaseng.2025.109576

Fujiang Zeng , Weilin Zhai , Yaman Jing , Rongzhu Zhang

To address the challenges of nonlinear intensity distortion (NID) and non-rigid shapes in dynamic fluid image registration, this paper proposes a multi-scale registration method based on multiple frame sequences (TSM-FIR). In this paper, a simplified Co-occurrence filtering (CoF) image pyramid is constructed to resist scale variations, and the phase coherence (PC) analysis is introduced at each pyramid layer to preserve the edge topology of the fluid field. A dual-path masked feature extraction (DMFE) strategy is designed to fuse feature points in multi-frame differential PC maps and original PC maps to enhance feature stability through spatio-temporal continuity constraints. A three-dimensional dynamic descriptor, TIFT, is proposed to capture the time-varying information of the fluid field through a point-sampling model in a multi-frame spatio-temporal cube. The experimental validation of the flame object shows that TSM-FIR performs stably with different flow rates and heterogeneous flames, with an average NCM of 55.368 and an average RMSE of 1.895 pixels, which is significantly better than that of the ablation group. It effectively solves the problem of registering the dynamic flow field without fixed morphology and provides a reliable technical basis for multimodal flow field analysis.

针对动态流体图像配准中存在的非线性强度失真（NID）和非刚性形状问题，提出了一种基于多帧序列的多尺度配准方法。本文构造了一个简化的共发生滤波（CoF）图像金字塔以抵抗尺度变化，并在每个金字塔层引入相相干（PC）分析以保持流场的边缘拓扑结构。设计了一种双路径掩蔽特征提取（DMFE）策略，将多帧差分PC地图与原始PC地图的特征点融合，通过时空连续性约束增强特征的稳定性。提出了一种三维动态描述符TIFT，通过点采样模型在多帧时空立方体中捕捉流体场的时变信息。火焰对象的实验验证表明，TSM-FIR在不同流量和非均质火焰下表现稳定，平均NCM为55.368，平均RMSE为1.895像素，明显优于烧蚀组。该方法有效地解决了无固定形态的动态流场注册问题，为多模态流场分析提供了可靠的技术依据。

{"title":"Multi-frame sequence-based multi-scale registration for dynamic fluid images","authors":"Fujiang Zeng , Weilin Zhai , Yaman Jing , Rongzhu Zhang","doi":"10.1016/j.optlaseng.2025.109576","DOIUrl":"10.1016/j.optlaseng.2025.109576","url":null,"abstract":"<div><div>To address the challenges of nonlinear intensity distortion (NID) and non-rigid shapes in dynamic fluid image registration, this paper proposes a multi-scale registration method based on multiple frame sequences (TSM-FIR). In this paper, a simplified Co-occurrence filtering (CoF) image pyramid is constructed to resist scale variations, and the phase coherence (PC) analysis is introduced at each pyramid layer to preserve the edge topology of the fluid field. A dual-path masked feature extraction (DMFE) strategy is designed to fuse feature points in multi-frame differential PC maps and original PC maps to enhance feature stability through spatio-temporal continuity constraints. A three-dimensional dynamic descriptor, TIFT, is proposed to capture the time-varying information of the fluid field through a point-sampling model in a multi-frame spatio-temporal cube. The experimental validation of the flame object shows that TSM-FIR performs stably with different flow rates and heterogeneous flames, with an average NCM of 55.368 and an average RMSE of 1.895 pixels, which is significantly better than that of the ablation group. It effectively solves the problem of registering the dynamic flow field without fixed morphology and provides a reliable technical basis for multimodal flow field analysis.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109576"},"PeriodicalIF":3.7,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Accelerated hologram generation using separable Fraunhofer diffraction 利用可分离夫琅和费衍射加速全息图生成

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-27 DOI: 10.1016/j.optlaseng.2025.109566

Noritomo Okamoto , Shota Osada , Fan Wang , Tomoyoshi Ito , Tomoyoshi Shimobaba

Holographic displays require a significant amount of time to generate holograms. One of the methods used to solve this problem is the fast computation of Fresnel holograms using separable convolution. In this study, we propose a separable Fraunhofer hologram that can compute holograms even faster. In addition, we derive recurrence formulas suited to the proposed methods. We compare the computational speed and image quality of the proposed methods with those of the conventional method. Although a dedicated computer using a field-programmable gate array (FPGA) has been developed, the conventional algorithm has a problem of increasing the complexity of computational circuitry. In this study, we demonstrate that the proposed method can simplify a computational circuit.

全息显示器需要大量的时间来生成全息图。解决这一问题的方法之一是利用可分离卷积快速计算菲涅耳全息图。在这项研究中，我们提出了一种可分离的弗劳恩霍夫全息图，可以更快地计算全息图。此外，我们还推导出适合所提出方法的递推公式。我们将所提方法的计算速度和图像质量与传统方法进行了比较。尽管使用现场可编程门阵列（FPGA）的专用计算机已经开发出来，但传统算法存在计算电路复杂性增加的问题。在这项研究中，我们证明了该方法可以简化计算电路。

引用次数: 0

FP-LD strong feedback fiber-optic stress system for high-voltage transmission towers 高压输电塔FP-LD强反馈光纤应力系统

IF 3.7 2区工程技术 Q2 OPTICS

Optics and Lasers in Engineering

Pub Date : 2025-12-27 DOI: 10.1016/j.optlaseng.2025.109579

Fang Wang , Yuchang Wen , Guoqing Shangguan , Shuangshuang Han , Xinyi Zhao , Yanzhong Yuan , Hualei Shen , Yufang Liu

We design a fiber-optic stress system based on strong optical feedback from a Fabry–Perot laser diode (FP-LD). It monitors stress changes induced by tilting of high-voltage transmission towers. The system comprises an external resonant cavity formed between the rear facet of the FP-LD and the end face of a single-mode fiber (SMF). The effective optical range length of the resonant cavity varies, which leads to a shift in the beat-frequency signal (BFS) of the multiple longitudinal modes. In the 0–80 N applied stress loading experiments of the sensing fiber, the stress variation is linearly related to the frequency shift of the BFS. The system sensitivity increased with increasing sensing resonator length. Specifically, for resonator lengths of 3.0 m, 5.1 m and 7.2 m, the measured stress sensitivities were −6.70 kHz·N⁻¹, −9.95 kHz·N⁻¹ and −13.67 kHz·N⁻¹, respectively. Applying a multilayer perceptron (MLP) neural network reduced the error between the stress values predicted from the frequency-shift monitoring software and the true stress measurements. This method improves the detection accuracy of the system and its MLP model has an accuracy of 96.88% in the test set.

本文设计了一种基于Fabry-Perot激光二极管（FP-LD）强光反馈的光纤应力系统。它监测高压输电塔倾斜引起的应力变化。该系统包括在FP-LD的后面和单模光纤（SMF）的端面之间形成的外部谐振腔。谐振腔的有效光学范围长度发生变化，导致多个纵向模式的热频信号发生移位。在传感光纤0 ~ 80 N的外加应力加载实验中，应力变化与BFS的频移呈线性关系。系统灵敏度随传感谐振腔长度的增加而增加。具体来说，对于长度为3.0 m， 5.1 m和7.2 m的谐振器，测量的应力灵敏度分别为- 6.70 kHz·N⁻¹，- 9.95 kHz·N⁻¹和- 13.67 kHz·N⁻¹。应用多层感知器（MLP）神经网络减少了频移监测软件预测的应力值与真实应力测量值之间的误差。该方法提高了系统的检测精度，其MLP模型在测试集中准确率达到96.88%。

{"title":"FP-LD strong feedback fiber-optic stress system for high-voltage transmission towers","authors":"Fang Wang , Yuchang Wen , Guoqing Shangguan , Shuangshuang Han , Xinyi Zhao , Yanzhong Yuan , Hualei Shen , Yufang Liu","doi":"10.1016/j.optlaseng.2025.109579","DOIUrl":"10.1016/j.optlaseng.2025.109579","url":null,"abstract":"<div><div>We design a fiber-optic stress system based on strong optical feedback from a Fabry–Perot laser diode (FP-LD). It monitors stress changes induced by tilting of high-voltage transmission towers. The system comprises an external resonant cavity formed between the rear facet of the FP-LD and the end face of a single-mode fiber (SMF). The effective optical range length of the resonant cavity varies, which leads to a shift in the beat-frequency signal (BFS) of the multiple longitudinal modes. In the 0–80 N applied stress loading experiments of the sensing fiber, the stress variation is linearly related to the frequency shift of the BFS. The system sensitivity increased with increasing sensing resonator length. Specifically, for resonator lengths of 3.0 m, 5.1 m and 7.2 m, the measured stress sensitivities were −6.70 kHz·N⁻¹, −9.95 kHz·N⁻¹ and −13.67 kHz·N⁻¹, respectively. Applying a multilayer perceptron (MLP) neural network reduced the error between the stress values predicted from the frequency-shift monitoring software and the true stress measurements. This method improves the detection accuracy of the system and its MLP model has an accuracy of 96.88% in the test set.</div></div>","PeriodicalId":49719,"journal":{"name":"Optics and Lasers in Engineering","volume":"199 ","pages":"Article 109579"},"PeriodicalIF":3.7,"publicationDate":"2025-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145842190","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0