最新刊期
卷 34 ， 期 2 ， 2026
- “风云三号F星紫外临边探测仪研究取得进展，专家推导观测几何指向模型，研发配置参数模式，为痕量气体反演提供解决方案。”摘要：Accurate assessment and correction of on-orbit pointing errors and tangent height deviations of limb sounders are essential for achieving high vertical accuracy in global trace retrievals. For the newly launched Ozone Monitoring Suite - Limb （OMS-L） onboard Fengyun-3F， the geometric pointing model of the scanning mirror was established based on the precisely measured instrument coordinate system. The tangent height calculation model was constructed using an equidistant virtual ellipsoid. Eight configuration parameter modes were designed to analyze the influence of the earth model， orbital and attitude information， installation parameters， and their combined effects on the theoretically designed tangent heights. During on-orbit testing， the systematic tangent height deviation was identified by knee-point method. To correct this deviation， the pitch installation parameters were adjusted， and the pointing stability of the instrument was statistically analyzed over a period of 12 months. The results show that the tangent height correction accuracy is better than 0.4 km， and the annual inter-annual variation of the instrument vertical （along-track） pointing is less than 0.01°. For cross-track pointing accuracy， image matching was performed using reference images acquired from the same platform and spectral band of the earth observation instrument. The results show that the cross-track pointing deviation of OMS-L reach 0.1°.The on-orbit pointing accuracy of OMS-L met the requirements for trace gas inversion and retrieval.关键词：atmospheric tangent detection;Ultraviolet;FengYun-3F;observation geometry;pointing accuracy;image matching0|0|0更新时间：2026-01-29
- “在光学设计领域，专家提出了基于节点像差理论的降敏设计方法，为手机镜头性能优化提供解决方案。”摘要：In the optical design process， the optimization of optical system performance and tolerance analysis are often treated separately. A system with excellent theoretical performance may have excessively tight tolerance. Therefore， the desensitization of the optical system is crucial during the system design process. The even aspheric surfaces are commonly used in the mobile phone lens to reduce aberrations and total track length. However， the introduction of high-order terms can lead to significant image quality degradation caused by the misalignments. A desensitization design method based on nodal aberration theory was proposed in this paper to desensitize mobile phone lenses with even aspheric surfaces . First， based on nodal aberration theory， the variations in Zernike coefficients for third-order coma and astigmatism were analyzed for the conic surface base of the misaligned even aspheric surfaces. Then， in order to calculate the variation in wavefront error introduced by the higher-order terms， a transformation of the pupil coordinates was used to obtain the pupil map of the optical surface under misalignments. Based on the coordinate transformation in the pupil map， the variation in wavefront error was then calculated. Finally， an as-built performance model suitable for optical systems with even aspheric surfaces was established， which was then used to design a 4-element mobile phone lens. The tolerance analysis showed that after 30 minutes of optimization and under 5 μm decenter and 5′ tilt tolerance， the mean RMS wavefront error of the mobile phone lens was reduced by 33% compared to the default design method， which verifies the effectiveness of the proposed design method in the paper.关键词：nodal aberration theory;desensitization design;aspheric surface;mobile phone lens0|0|0更新时间：2026-01-29
- “在电力设备巡检领域，研究人员设计了基于热电制冷器的主动温控模块，有效提升了双通道红外探测系统在SF6气体泄漏检测中的稳定性和性能。”摘要：To address performance degradation caused by detector temperature fluctuations in a dual-channel infrared imaging system for sulfur hexafluoride （SF6） gas leak detection， this study developed an active temperature control module based on a Thermoelectric Cooler （TEC） to enhance its stability. The optimal spatial layout of the TEC was determined by combining experimental measurements with thermal field simulations， and a closed-loop control system governed by a PID algorithm was implemented. Comparative experiments demonstrated that the integrated control module significantly enhances the system performance. Regarding imaging quality， the steady-state standard deviation of the grayscale value is reduced by 40.87%， the linearity error is decreased from 7.253 4% to 0.660 4%， and the Noise Equivalent Temperature Difference （NETD） is improved by 19.35%， from 14.379 mK to 11.597 mK. In the SF6 gas detection application， the average relative contrast， signal-to-noise ratio， and gas pixel ratio are increased by 123%， 56%， and 77%， respectively. This enhancement method effectively suppresses the adverse effects of temperature fluctuations， substantially improving SF6 gas detection capability while enhancing the fundamental imaging quality， and validating a feasible technical pathway for developing high-performance， low-cost portable gas leak detection instruments.关键词：infrared imaging;Sulfur hexafluoride （SF6）;uncooled detector;temperature control;thermoelectric cooling;dual-channel system0|0|0更新时间：2026-01-29
Modern Applied Optics
- “在压电定位台高精度跟踪控制领域，研究人员设计了分数阶终端滑模自抗扰控制器，有效提高了系统跟踪控制性能，削弱了抖振现象，对模型不确定性及负载扰动具有强鲁棒性。”摘要：To achieve high-precision tracking of the piezoelectric positioning stage， a fractional-order terminal sliding mode active disturbance rejection controller was developed to address hysteresis， model uncertainties， and load disturbances. By integrating fractional-order theory， terminal sliding mode， and active disturbance rejection， the controller ensured finite-time error convergence. A continuous and differentiable reaching law based on an improved fal nonlinear function was introduced to suppress chattering. Experimental results show that for a 10 μm sinusoidal trajectory under a 200 g load， the controller attains a root mean square error of 0.024 6 μm， improving accuracy by 76.66%， 48.00%， and 43.32% compared with PID， integer-order terminal sliding mode， and neural-network-based integer-order terminal sliding mode control， respectively. Under a 40 Hz reference signal， it reduces the average tracking error to 0.186 9 μm. The controller maintains excellent tracking across varying loads， effectively suppresses chattering， and exhibits strong robustness to model uncertainties and load disturbances.关键词：piezoelectric positioning stage;fractional order;sliding mode control;Fal function;extended state observer0|0|0更新时间：2026-01-29
- “在激光通信领域，专家设计了压电陶瓷驱动快速反射镜控制算法，有效提升跟踪性能和适应能力。”摘要：To meet the demanding requirements of high-precision alignment and tracking in laser communication systems operating under complex environmental conditions， this paper presented a novel control algorithm for a piezoelectric-actuated fast steering mirror （FSM）. A Hammerstein-structured nonlinear model incorporating a Prandtl-Ishlinskii hysteresis representation was established for the piezoelectric actuator， and system linearization was achieved through a geometrically derived inverse model employed as a feedforward compensator. For the linearized system， a hybrid control strategy integrating radial basis function （RBF） neural networks with sliding mode control was designed， where the RBF network approximated unmodeled nonlinear parameters within the nonsingular fast terminal sliding mode control framework to enhance robustness while maintaining tracking precision. Experimental validation on a servo platform confirms the algorithm's effectiveness in tracking composite-frequency trajectories with a root mean square error below 2.7 μrad. Under 10% model mismatch conditions， the method maintains a root mean square error under 7.3 μrad when tracking 50 Hz sinusoidal signals， representing performance improvements of approximately 40.16% and 75.9% over conventional nonsingular fast terminal sliding mode control and active disturbance rejection control， respectively， while demonstrating strong robustness against parameter perturbations.关键词：laser communication;piezoelectric actuation;fast steering mirror;radial basis function neural network;sliding mode control0|0|0更新时间：2026-01-29
- “本报道介绍了航空动态光学成像系统中音圈电机驱动反射镜的高精度控制策略，专家建立了自适应滑模有限时间控制体系，为解决扰动抑制与跟踪控制问题提供解决方案。”摘要：To enhance the control precision of aerial dynamic optical imaging systems， this study addresses the high-precision control problem of Voice Coil Motor-Driven Mirrors （VCMM） by proposing an integrated control strategy based on an adaptive sliding mode finite-time control. This strategy enables precise command tracking under conditions of external disturbances and model uncertainties. First， a generalized form of the system dynamics model was established based on the electromechanical characteristics of the VCMM. Second， a nonlinear proportional-integral-derivative type sliding surface was constructed to improve the convergence speed of the control system and suppress steady-state errors. Based on this， following a hierarchical anti-disturbance framework， feedback control terms related to model parameters and errors， sliding mode switching control terms， and adaptive control terms were designed. The finite-time convergence characteristics of the closed-loop system were rigorously demonstrated using Lyapunov stability theory. Finally， the effectiveness of the proposed composite control strategy was validated through comparative experiments. Experimental results showed that under a sinusoidal command with an amplitude of 180 arcseconds and a frequency of 1 Hz， and in the presence of periodic bounded disturbances， the system achieved tracking errors with root mean square and peak-to-peak values of only 0.28 arcseconds and 4.04 arcseconds， respectively， significantly outperforming conventional sliding mode control. The proposed control algorithm not only maintains excellent tracking performance but also demonstrates superior disturbance suppression capabilities. This study provides an effective solution for addressing disturbance rejection and tracking control in high-precision aerial optoelectronic imaging systems.关键词：voice coil mirror-driven mirror;aerial dynamic optical imaging systems;sliding mode control;Adaptive control;finite-time convergence0|0|0更新时间：2026-01-29
- “在微纳定位与精密制造领域，针对压电作动器迟滞非线性问题，研究者提出了多层神经网络改进型P-I建模方法，有效提升了系统控制精度。”摘要：Piezoelectric actuators （PEA） are widely used for micro-nano-positioning and precision manufacturing due to their high resolution and rapid response. Inherent hysteresis nonlinearity affects control performance and restricts high-accuracy applications. To overcome the limitations of the classical Prandtl-Ishlinskii （P-I） model in representing complex nonlinear hysteresis phenomena， a multilayer neural network-enhanced P-I modeling approach was proposed. The method used a neural network to dynamically map the weights of Play operators while ensuring that the model remained invertible and physically interpretable. Bayesian regularization was adopted during training to improve the ability to fit nonlinear systems and enhance generalization. Based on the improved model， an inverse-model-based feedforward controller was designed and validated in real-time experiments. Experimental results show that the proposed feedforward compensation reduces the normalized RMSE to 0.65%， 0.76%， and 1.82% under triangular， sinusoidal， and hybrid inputs， significantly outperforming the classical and its polynomial variants. The method exhibits strong robustness across diverse input conditions and demonstrates good engineering applicability in complex hysteresis modeling and high-precision control.关键词：piezoelectric actuators;nonlinear modeling;multilayer neural network;hysteresis compensation0|0|0更新时间：2026-01-29
Micro/Nano Technology and Fine Mechanics
- “在航拍图像语义分割领域，专家提出了跨层级交互与方位感知网络，有效提升了分割精度和泛化能力。”摘要：To address the issues of single-scale feature extraction， detail loss， and blurred boundaries in aerial image semantic segmentation， this paper proposed an aerial image semantic segmentation network with cross-level interaction and orientation awareness. A position awareness module was constructed through a direction-decoupled attention strategy to enhance the model's ability to process spatial directional information； a cross-level interaction module was designed for inter-channel feature interaction and fusion to improve spatial perception， while a channel-spatial attention mechanism was used to enhance feature extraction capabilities and alleviate detail blurring issues in complex scenes； finally， a lightweight design was implemented for the segmentation head， removing redundant operations to reduce computational load while ensuring segmentation performance. Experimental results indicate that the proposed network achieves a 1.7% and 1.3% improvement in mean intersection over union on the UAVid and Aeroscapes datasets， respectively， compared to the baseline model SegFormer， demonstrating the network's effectiveness in semantic segmentation under complex conditions such as aerial images. The segmentation accuracy of the Human category improved by 1.8% compared to the baseline model， demonstrating that the network proposed in this paper performs excellently in small object segmentation. Compared with several mainstream networks， the method proposed in this paper achieves the highest segmentation accuracy on both datasets， showing superior generalization capability.关键词：inter-level interaction;semantic segmentation;aerial imagery;lightweighting;direction decoupling0|0|0更新时间：2026-01-29
- “在高光谱图像分类领域，专家构建了高斯交叉特征融合模型，验证了其在植物、人工地物和异构混合场景中的分类优势，为该领域研究提供了新方向。”摘要：To address the persistent imbalance between local and global representations in hyperspectral image （HSI） classification， a Gaussian cross-feature fusion method （GCFFM） was proposed to explicitly balance the contributions of heterogeneous cues during feature integration. Interactive couplings between local and global feature streams were established via element-wise multiplication， enabling direct cross-modulation of complementary information. Building on this interaction， a Gaussian cross-attention fusion algorithm was devised to infer intrinsic relationships from the similarity of projected feature vectors. Fusion weights were parameterized by a Gaussian function， and the associated key parameters were optimized end-to-end so that the model could dynamically reweight local and global features according to scene content， thereby maintaining a balanced contribution throughout the fusion process. This design targeted common failure modes in HSI classification-such as over-emphasis on fine textures at the expense of scene context or， conversely， dominance of broad context that blurred class boundaries. Comprehensive experiments were conducted on four public benchmarks-Indian Pines， Pavia University， Salinas， and LongKou-against eleven mainstream baselines， including DCSST， SMESC， and ViT-cov. Performance was evaluated by using Overall Accuracy （OA）， Average Accuracy （AA）， and the Kappa coefficient. The proposed method achieves the highest OA on each dataset： 96.67% on Indian Pines， 98.16% on Pavia University， 98.81% on Salinas， and 97.09% on LongKou. These gains are accompanied by consistent improvements in AA and Kappa， indicating that the model not only raises overall correctness but also enhances per-class reliability and agreement beyond chance. Overall， GCFFM delivers consistently superior classification across agricultural， urban， and heterogeneous mixed scenes， which demonstrates strong generalization and robustness. By coupling element-wise cross-feature interactions with Gaussian-parameterized attention， the method provides a principled and effective solution to balancing local detail and global context in HSI classification.关键词：hyperspectral image classification;attention mechanism;Gaussian function;feature extraction;feature fusion0|0|0更新时间：2026-01-29
- “最新研究突破工业异常检测难题，提出融合高频残差和多尺度注意力的检测网络，显著提升小缺陷检测精度和异常分割效果。”摘要：Aiming at the problems of low detection accuracy for small-sized defects， weak multi-scale feature extraction ability and low anomaly segmentation accuracy of existing industrial anomaly detection algorithms， an industrial anomaly detection network combining high-frequency residual guidance and multi-scale attention feature fusion was proposed. Firstly， aiming at the problem of high-frequency detail loss caused by traditional full-frequency processing， a frequency-domain separation strategy was designed. Gaussian kernel filtering was utilized to extract high-frequency residual features， enhancing the network's detection ability for minor anomalies. Secondly， aiming at the problems of insufficient representation ability of complex textures and low discrimination between anomalies and backgrounds in conventional convolutional networks， a globally enhanced multi-scale attention module GEMA is embedded in the encoder stage of the discriminative network. It captures multi-scale local information in the horizontal and vertical directions through parallel dual-path， enhancing the salient features at different spatial positions. Improve the feature discriminability in complex texture backgrounds； Finally， in the decoder stage of the discriminant network， the coordinate attention module CoordAtt is integrated. By decomposing the coordinate axes and dynamically modulating the feature weights， precise spatial positioning of abnormal areas is achieved. Experiments show that on the MVTec AD public dataset， the average AUROC at the image level of the improved model is 98.6%， and the average AUROC and AP at the pixel level are 97.6% and 73.2% respectively， effectively improving the effect of industrial anomaly detection.关键词：industrial anomaly detection;high-frequency component guidance;multi-scale spatial perception;attention mechanism0|0|0更新时间：2026-01-29
- “在矿石粒度检测领域，研究者提出了一种小样本目标检测算法，有效提升了感知性能，为智能矿山实时检测提供了技术方案。”摘要：To address the high computational complexity， limited feature robustness， and constrained classifier performance of conventional object detection methods in ore particle size detection， a few-shot object detection approach was proposed to reduce annotation cost and improve generalization under data-scarce conditions. The proposed method was built upon the CenterNet2 framework and employed a lightweight VoVNet as the backbone to ensure detection efficiency. A parallel dual-attention feature fusion module was designed as the core component. Specifically， a channel cross-attention module was introduced to recalibrate channel-wise feature responses， while a spatial group-attention module emphasized discriminative target regions. The coordinated operation of the two modules enhanced the fusion of task-relevant features and provided effective guidance for query image detection in few-shot scenarios. Experimental results on an ore dataset show that the proposed model achieved an average precision （AP） of 55.2%， with AP50 and AP75 reaching 78.5% and 66.9%， respectively. The inference speed reached 57 frames per second（FPS）， while the attention module required only 16.1 M parameters， indicating a favorable trade-off between accuracy and efficiency. Experimental results demonstrate that the proposed method effectively enhances the perception performance of few-shot ore particle size detection. Moreover， it possesses high potential for edge deployment， providing a reliable technical solution for real-time detection challenges in smart mines under computation-constrained conditions.关键词：computer vision;few-shot object detection;lightweight;ore images;real-time0|0|0更新时间：2026-01-29
- “在水下图像增强领域，研究者提出了改进U-Net的全局特征融合网络，有效恢复颜色偏差和减少细节模糊。”摘要：To address the color deviation and detail blur of underwater images caused by scattering and attenuation when light propagated in the underwater environment， an improved U-Net global feature fusion underwater image enhancement network was proposed. Firstly， a multi-residual convolution module was designed in the encoder and decoder to fuse the feature information hierarchically to reduce the loss of detail information. Secondly， the channel attention module was introduced into the decoder to weight the channels to alleviate the problem of different degrees of channel degradation. Finally， a convolution-permuted self-attention module was designed in the decoder to fuse the global information and promote the network-guided image reconstruction. The proposed method was tested on UIEB dataset， and finally achieved 23.42， 0.900 5 and 0.138 5 on PSNR， SSIM and LPIPS， respectively. The results on LSUI dataset were 29.35， 0.938 2 and 0.088 0 on PSNR， SSIM and LPIPS， respectively. Compared with other commonly used underwater image enhancement methods on several public underwater image datasets， the experimental results show that the proposed method has good effect in restoring color deviation and reducing detail blur， which proves its effectiveness and feasibility.关键词：underwater image enhancement;deep learning;feature fusion;attention mechanism;convolutional neural network0|0|0更新时间：2026-01-29
- “在路面裂缝检测领域，研究者提出了一种自适应特征的轻量化检测方法，显著提升了检测精度和效率，为自动化道路维护提供技术方案。”摘要：To address the issues of complex crack morphologies， environmental interference， and the imbalance between detection accuracy and model lightweight requirements in road surface inspection， this paper proposed a lightweight road crack detection method with adaptive feature extraction. First， a Crack Efficient Attention （CEA） module was designed based on the slender shape and large span characteristics of cracks， compressing feature dimensions to capture long-distance spatial dependencies. Second， a Dynamic Sampling Feature Pyramid Network （DSFPN） was constructed for adaptive sampling and target feature extraction， enhancing representation capability for heterogeneous crack features. Third， the HGNet_GS lightweight backbone network was improved， and a CEA Group Head （CGHead） was proposed， significantly reducing computational redundancy； the PIoU （Powerful IoU） loss function was adopted to solve anchor box expansion problems and improve convergence speed for small models. Additionally， a civilian road defect dataset containing 2 985 images under various lighting conditions was established to validate model generalization. Experimental results show that compared with the baseline YOLOv8n model， the proposed method reduces parameters and computational cost by 50% and 52%， respectively； on the self-built dataset， mAP50 and mAP95 increase by 5.4% and 4.1%； on the public RDD2022 dataset， these metrics improve by 2.1% and 1.5%. The model has been deployed on edge devices and verified through engineering tests， demonstrating its capability to meet practical requirements for lightweight road crack detection and providing a technical solution for automated road maintenance systems.关键词：pavement crack detection;attention mechanism;lightweight;dynamic pyramid;YOLOv80|0|0更新时间：2026-01-29
Information Sciences
- Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China Postal code：130033
- Tel：0431-86176855 Email：gxjmgc@ciomp.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0