The directional polarimetric camera (DPC) aboard the Hyperspectral Observation Satellite (GF-5-02) has been launched successfully in 7 September 2021. The wide-field relative radiometric calibration for DPC is to minimize pixel-to-pixel response variations to ensure spatial radiometric uniformity. It effectively reduces the errors present in the radiometric data collected from multi-angle observations of a target. This process makes it easier to accurately describe the distribution pattern of the target's reflected radiometric information, thereby enhancing the precision of target property retrieval. Therefore, the relative radiometric calibration is of great significance for wide-field-of-view sensors. Due to the absence of onboard calibrator and the constraints imposed by traditional calibration sites, accurately correcting for non-uniformity in irradiance response across a vast number of DPC pixels continues to pose a significant challenge. Therefore, we propose a novel approach for relative radiometric calibration of wide field-of-view array sensors utilizing large-area high-altitude thick ice clouds as reference targets. The new approach comprehensively considers the requirements for cross-calibration across the spectral range, swath width, and observational synchronicity between two sensors. It employs the data from the Advanced Baseline Imager onboard the Geostationary Operational Environmental Satellite (GOES/ABI) to monitor the relative radiometric properties of the DPC at the polarized bands of 490 nm, 670 nm and 865 nm. The measurement results, covering the period from November 2021 and November 2023, indicate that the annual relative radiometric changes are less than 6 % (4.45 % at 490 nm, 5.07 % at 670 nm, 5.23 % at 865 nm). The attenuation in the edge field of view is less than that in the central field of view. This method verifies the feasibility of monitoring the relative radiometric performance of wide field-of-view sensors using stationary meteorological satellites, providing critical support for the quantitative application of multi-angle observation sensors and offering a new reference technology for radiometric calibration of wide field-of-view and large swath-width sensors.

Cloud segmentation is a fundamental step in remote sensing analysis, and its accuracy directly affects downstream processing. Vision foundation models pretrained on large-scale datasets exhibit strong transfer potential. However, due to the large domain gap between natural and remote sensing images, including differences in viewing geometry, spatial resolution, and object scale, directly transferring such models to cloud segmentation often yields suboptimal performance. To address this, we propose DinoCloudNet, built on a frozen DINOv3 backbone, which introduces two key modules: The Efficient Channel Attention Spatial Perception Module (ECA-SPM) and the Multi-Scale Global-Local Feature Fusion Adapter (MGFF Adapter). DinoCloudNet adopts a dual-branch design: the frozen backbone extracts global semantics, while a parallel ECA-SPM branch captures edge-aware, multiscale local spatial details. The MGFF Adapter then adaptively aligns and fuses multiscale information from both branches, injecting the fused features into selected layers of the backbone. This parameter-efficient approach enables effective adaptation to the cloud task while preserving the strong generalization of the frozen backbone, leading to improved accuracy. Experiments show that DinoCloudNet achieves highly competitive performance on multiple cloud segmentation benchmarks, outperforming most existing methods. This performance is consistent across diverse satellite platforms, sensor families, and land cover conditions, demonstrating the effectiveness and generalization ability of the proposed approach.

Polarization information can effectively characterize the scattering properties of targets, holding significant importance in enhancing the retrieval accuracy of earth observation products. During its orbital operation, the first multi-angular polarization sensor (POLDER) collected extensive measurement data incorporated multiple angles and polarization information. The BPDF model and its derived DOLP model, obtained based on POLDER data, have been widely adopted in the area of polarization retrieval. Due to the limited availability of on-orbit multi-angular polarization sensors, the validation of the POLDER/DOLP models using alternative sensors has remained constrained. In this research, we utilized two multi-angular sensors operating at different orbital altitudes and incorporated desert site observation data (characterized by high stability and spatial uniformity) to perform preliminary validation of the POLDER/DOLP models. The validated results indicate that: For the desert samples utilized, the POLDER/DOLP models exhibit good accuracy and similarity, with minimal seasonal variation. This validates the models’ effectiveness across various seasons. The measured data of two sensors and the POLDER/DOLP models exhibit a strong linear correlation, with Root Mean Square Error (RMSE) of 0.35% and 0.64%, and Pearson Correlation coefficient of 0.9781 and 0.9324, respectively. These results provide preliminary evidence supporting the generalizability of the POLDER/DOLP models to other polarization sensors. When referred to the POLDER/DOLP models, the polarization measurement residuals for both sensors consistently remain within 0.02, demonstrating their stable performance. Collectively, these results establish a foundational benchmark for further comprehensive validation and enhancement of polarization models, reinforcing the preliminary confirmation of the broad applicability of the POLDER/DOLP models.

Overlay is a key determinant of lithography quality, and IBO is one of the mainstream techniques for overlay metrology. A crucial challenge in IBO is minimizing tool induced shift (TIS). We analyze, via simulation and experiment, how chromatic aberration and wedge induced dispersion (WID) affect TIS. Incorporating a realistic field of view decenter, we identify TIS-insensitive regions of lateral chromatic aberration to constrain optical design and show that WID couples to target layer thickness and focus accuracy. We derive compensation for a double-wedge prism corrector and demonstrate WID reduction from hundreds of nanometers to less than 3 nm. We also propose a fast WID testing method that mitigates temperature drift, achieving about 2 nm (3σ) repeatability, guiding advanced node IBO.

Cloud Droplet Size Distributions (CDSD) play a crucial role in cloud microphysics and cloud-radiationprecipitation interactions. Assuming a gamma distribution, CDSD can be parameterized by the Cloud Effective Radius (CER) and the Cloud Effective Variance (CEV), representing the characteristic droplet size and the width of the size distribution, respectively. Multi-Angular Polarimetric (MAP) measurements currently provide the richest information content for simultaneously retrieving both the CER (r(eff)) and CEV (upsilon(eff)) of liquid clouds, provided that the multi-angle observations adequately sample the cloud-bow angular region. However, the strict requirements of MAP retrieval methods on the angular distribution sampling and cloud heterogeneity have hindered the realization of high spatial-resolution global CDSD retrievals. Consequently, to guarantee sufficient sampling of the rainbow angular domain, earlier MAP inversions have largely depended on aggressive spatial aggregation, such as the retrieval at similar to 150 km spatial-resolution for POLDER. Here, we exploit MAP measurements from the Directional Polarization Camera (DPC) onboard GaoFen-5 (02) (GF-5 (02)) satellite to investigate the key factors that constrain CDSD retrievals. Leveraging pre-inversion MAP Information Content (IC) analysis and neighboring-pixel weighted contributions during inversion, we develop a dedicated high spatial-resolution CDSD retrieval method based on dynamic pixel-aggregation tailored to global DPC/GF-5 (02) observations. IC simulations indicate that CER retrieval is more robust than CEV. The resulting products are evaluated against MODIS r(eff) product to quantify retrieval performance. Consistent with POLDER-based studies, our DPC retrievals of r(eff) show a global mean bias of approximately 4.8 mu m smaller relative to the MODIS product from dual-channel approach with fixed upsilon(eff) in January 2024. Notably, the CDSD retrieval resolution improves from similar to 150 km for the POLDER-based product to 3.3 km for DPC, owing to both the higher native spatial resolution of DPC and the proposed dynamic pixel-aggregation approach. Furthermore, under sufficient rainbow-angle (135(degrees)-165(degrees)) coverage, successful retrievals over ocean and land require aggregation scales no larger than 3 x 3 native DPC pixels for 68.84% and 72.09% of pixels, respectively. This result suggests that the pixel-aggregation strategies commonly adopted in current MAP retrievals may be overly conservative. Overall, this method represents an important step toward stable, high spatial-resolution global CDSD products from satellite MAP measurements and provides valuable constraints for future cloud property retrievals and climate-related applications.

The Directional Polarization Camera (DPC) is a Chinese Earth observation satellite sensor carried on Gaofen-5B satellite, which has observation capacity of multi-angle, multi-spectral, and polarization. Due to its wide field of view (±50°) and short revisit period (2 days), the DPC is capable of monitoring global changes in aerosol optical depth (AOD). In this paper, we develop a DPC AOD retrieval algorithm over open ocean waters based on the black pixel assumption in the near-infrared band, namely at 865 nm. The optimization of the retrieval process is achieved by matching the multi-angle measured reflectance with the reflectance from the radiative transfer simulation. First, based on the so-called Ocean Successive Orders with Atmosphere—Advanced (OSOAA) radiative transfer model, the AOD lookup table is constructed by storing the simulated reflectance at the top of atmosphere (TOA) corresponding to various AODs. Then, the satellite’s observation geometry is used to interpolate the simulated reflectance. The retrieval scheme of this method is to minimize the residuals between the measured directional data and the simulated data in each direction. Specifically, a second-order polynomial is used to fit the cost function for each observation geometry based on the least squares method. By solving this polynomial, a set of AOD values for the pixel is obtained, and the median of this set is taken as the final AOD. Finally, we validate the effectiveness of the proposed algorithm by conducting a pixel-by-pixel comparison of the retrieval results of the DPC and the Moderate Resolution Imaging Spectroradiometer (MODIS) onboard the Terra satellite. The results show that the spatial distribution of DPC AOD is consistent with that of MODIS, with a coefficient correlation of determination (R2) of 0.726 and a root mean square error (RMSE) of 0.01. The results indicate that the proposed algorithm can provide effective technical support for AOD inversion of DPC over open ocean.

Cloud is essential components of the weather and climate system, influencing it by altering the radiation balance and participating in the hydrological cycle. Cloud Optical Thickness (COT) and Cloud Effective Radius (CER) are critical parameters describing cloud optical and microphysical properties. This study investigates the reflectance characteristics of the Particulate Observing Scanning Polarimeter onboard GaoFen-5(02) satellite (POSP/GF-5(02)) and proposes a reflectance correction method based on the Unified Linearized Vector Radiative Transfer Model (UNL-VRTM). Using the corrected reflectance, COT and CER are retrieved through an optimized dual-channel method (0.865-2.25 μm), with channel selection guided by sensitivity analysis. The retrieval results are further assessment with Moderate Resolution Imaging Spectroradiometer (MODIS) products. Results show that, under the current correction method, the original POSP reflectance used for cloud target is underestimated by 15.61%, 51.59%, and 44.38% at 0.865, 1.61, and 2.25 μm, respectively. Region case studies demonstrate that after reflectance correction, the fractions of retrievals within the predefined accuracy thresholds increase by an average of 61.83% for CER and 47.08% for COT. Further case analysis shows that retrieval accuracy varies with cloud field heterogeneity and surface complexity, with CER accuracy decreasing by 50.56% relative to homogeneous cloud field and COT accuracy decreasing by 13.23% relative to dark ocean surfaces. These results indicate that the proposed reflectance correction framework substantially enhances the accuracy of cloud property retrievals from POSP observations, providing improved capability for cloud remote sensing and supporting further studies of weather and climate processes.

Thin-film interference in photoresist stacks can become a significant source of uncertainty in lithographic focus metrology, particularly when high measurement stability is required. To evaluate this effect, a Fresnel-based multilayer reflection model is used to analyze the optical response of the resist stack and to guide the selection of dual-wavelength illumination. On this basis, a dual-wavelength optical triangulation system is developed for focus metrology in 350 nm lithography, with signal acquisition performed by a linear charge-coupled device (LCCD). Rather than improving precision by reducing detector pitch, the system employs a two-stage sub-pixel localization strategy in which template matching provides coarse spot localization and weighted centroid interpolation refines the final position within localized calculation windows, keeping the computational cost manageable. A covariance-based uncertainty analysis predicts a total root-mean-square uncertainty of 27.23 nm. Prototype experiments were performed on a bare silicon wafer to establish the intrinsic performance of the instrument before introducing process-dependent optical effects. Under these conditions, the system achieved a vertical resolution of 10 nm, a repeatability of 35 nm, and a stability of 13.16 nm. The additional uncertainty expected under resist-coated-wafer conditions was assessed separately through the thin-film model. These results verify the baseline capability of the proposed system and support the feasibility of the dual-wavelength strategy for focus metrology in 350 nm lithography.

Achieving high-precision overlay target center localization is critical for image-based overlay (IBO) metrology in advanced semiconductor manufacturing. This paper proposes a novel IBO target localization algorithm based on symmetry center matching. Leveraging the symmetry design of the IBO optical system as a physical prior, the algorithm reformulates center localization as a global correlation optimization problem. The grayscale projection profile of a single-sided edge is extracted, spatially mirrored, and used as a reference template for sliding correlation matching against the opposite edge. The symmetry center is then determined from the peak of the Pearson correlation coefficient curve. Simulation results demonstrate a center localization accuracy better than 0.00013 pixels (3σ), with repeatability precision remaining within 0.012 pixels (3σ) under stringent noise and blur conditions. Experimental validation yields object-space repeatability precision of 0.129 nm (3σ) and 0.144 nm (3σ) in the X and Y directions, respectively, surpassing the 0.32 nm measurement uncertainty requirement for advanced process nodes. The average single-frame processing time is approximately 0.07 s, demonstrating that the proposed algorithm simultaneously satisfies the demands of high precision and high throughput.

Achieving stable local focus-height measurement across different material surfaces is important for I-line-lithography-related inspection, where sub-micrometer height deviations can affect imaging quality, exposure uniformity, and subsequent autofocus performance. This study evaluates the local focus-height repeatability of a linear charge-coupled device (LCCD)-based focus metrology system under several I-line-lithography-related material-surface conditions. The prototype integrates fiber-coupled LED illumination, telecentric projection and imaging optics, reference marks, and a two-step localization procedure based on template matching and centroid estimation; the dual-wavelength source is treated as part of the fixed optical configuration. Tests were performed on silicon wafers, GaAs bright substrates, sapphire, infrared transmissive material, and SiC, covering different reflectivity levels and surface structures. The measured peak-to-valley repeatability was 35-37 nm for highly reflective samples and 40-54 nm for intermediate- or low-reflectivity and microstructured samples, all below the selected 70 nm conservative engineering criterion derived from the depth-of-focus estimate. These results indicate that the integrated LCCD measurement chain maintained stable local repeatability within the tested material-surface range, providing experimental support for further development of local focus metrology and precision optical inspection.

The Directional Polarimetric Camera (DPC) onboard the Terrestrial Ecosystem Carbon Inventory Satellite (TECIS) utilizes a staring imaging technique with a wide field of view (100 degrees along-track and 80 degrees cross-track), allowing it to capture polarization data from up to 17 observation directions. Due to the lack of onboard calibration and the limitations of traditional calibration sites, providing an accurate correction for non-uniformity effects in the irradiance response across all DPC pixels remains challenging. To address this issue, we propose a novel approach for the relative radiometric calibration which exploits vast snowfields in Antarctica. Crosscalibration is performed using data from the MODerate resolution Imaging Spectroradiometer (MODIS). Additionally, the Angular Distribution Model (ADM) of the Clouds and the Earth's Radiant Energy System (CERES) is applied to reduce anisotropic biases in the measured snow reflectance. This approach leverages the short revisit period of the polar-orbiting satellite over Antarctica to provide good synchronization. The relative radiometric performance of the DPC, monitored over the December 2022 - December 2023 period, indicates that the annual relative radiometric changes are +3.8% at 670 nm and +4.1% at 865 nm. The variations in the two bands were consistent in spatial distribution. This study confirms the feasibility of using Antarctic snow for relative radiometric calibration and provides a valuable reference for applications to similar wide-field-of-view, remote sensing instruments.

Satellite aerosol retrievals have usually relied on aerosol models derived from climatological and observational statistics (e.g., averaging or clustering), which are probably unable to adequately represent the continuity of atmospheric aerosol and meet assumption of linear mixing. This poses a challenge to improving retrieval accuracy and expanding the range of retrievable parameters. To address this issue, this study proposes an optimal aerosol model construction method based on the non-negative matrix factorization (NMF) approach. It allows reconstructing large amounts of data using a small set of basis aerosol models (BAMs) and explicitly enforces the externally linear mixing, making the resulting aerosol models directly compatible with forward radiative transfer calculations in satellite retrieval algorithm. This NMF-based BAMs are successfully applied to the Directional Polarimetric Camera (DPC) over land combined with the GRASP framework to retrieve aerosol optical and microphysical properties. Results show that the retrieved aerosol optical depth (AOD), & Aring;ngstrom exponent, and single scattering albedo (SSA) show good performance comparing with the AERONET data, with the correlation coefficients of 0.935, 0.71, and 0.644 and the mean biases of 0.015,-0.196, and 0.007, respectively. Error analysis further indicates that biases are most obvious in dust-dominated desert regions, and the accurate retrieval of SSA relies on the correct separation of aerosol particle size and high AOD conditions. Additionally, tests based on different samples indicate that the BAMs derived from the NMF exhibit good stability and representativeness, but their performance shows slight degradation under both very low and high AOD conditions. This is mainly due to the limited ability of a few aerosol models in linearly capturing the diversity and variability of real atmospheric aerosol conditions.

Multi-angle polarimetric (MAP) satellite observations provide information on aerosol optical and microphysical properties. In this study, we propose an effective transformer-based deep learning (DL) algorithm for aerosol retrieval using MAP observations, utilizing synergistic observations from the Directional Polarimetric Camera (DPC) and POSP (Particulate Observing Scanning Polarimeter) sensors of the polarization crossfire (PCF) sensor suite. The use of these two sensors overcomes limitations of traditional DL approaches that rely solely on sparse ground-based stations. The proposed algorithm involves two main steps: (1) a date base of over 41,000 high-confidence aerosol samples are is constructed using Aerosol Robotic Network (AERONET) data supplemented by selected high-quality-screening data retrieved from the Particulate Observing Scanning Polarimeter (POSP sensor) from using the Generalized Retrieval of Atmosphere and Surface Properties (GRASP) algorithm; (2) A transformer model is trained on DPC MAP measurements, to learn the nonlinear representation linking spaceborne measurements with targets for estimating aerosol optical depth (AOD), fine-mode AOD (FAOD), and coarse-mode AOD (CAOD) at 550 nm. Comparison of the aerosol parameters derived from application of the transformer model to DPC observational data to AERONET reference data observations, shows substantial improvement over models trained solely on AERONET data, with high Pearson correlation coefficients (R) of 0.855 (AOD), 0.812 (FAOD), and 0.793 (CAOD), and slopes of 0.837, 0.813, and 0.801. Comparison with POSP/GRASP aerosol products shows improved spatial coverage, especially over bright surfaces and during extreme aerosol events such as dust and forest fires. This scalable and transferable framework integrates optimal estimation (OE) interpretability with DL efficiency, achieving 3-4 orders of magnitude speedup over GRASP methods and offering a promising solution for next-generation MAP satellite missions.

Atmospheric correction of high-spatial-resolution (HSR) optical satellite imagery is strongly affected by the adjacency effect (AE). Conventional Atmospheric Point Spread Function (APSF)-based AE correction methods are often based on simple local averaging or distance-weighted background approximations. These are often insufficient for highly heterogeneous HSR scenes and become computationally expensive when the AE’s influence range spans far more pixels. To address these issues, this study proposes a fast AE correction algorithm with adaptive local surface constraints. The method first introduces a surface-atmosphere coupling correction based on effective reflectance. It then constructs downward- and upward-weighting kernels and incorporates local reflectance constraints into the estimation of environmental reflectance to better characterize AE intensity in HSR scenes. Since environmental reflectance estimation is retained in a kernel-weighted form, the infinite-domain integration is reformulated as a finite-window computation with truncation compensation and accelerated via fast Fourier transform (FFT) convolution, followed by a few iterations for reflectance retrieval. Validation with GaoFen-2 (GF-2) panchromatic imagery shows that, at an aerosol optical depth of about 0.4, the proposed method achieves the best performance among the compared methods, with a mean absolute error (MAE) below 0.006 relative to in situ measurements, sharpness and contrast increases of approximately 99.0% and 97.9%, respectively, and a National Imagery Interpretability Rating Scale (NIIRS) increase of more than 0.3. For a 1024×1024 image with a 501-pixel AE window diameter, the running time is below 4 s, substantially lower than that of previous APSF-based AE correction methods. The FFT implementation also avoids the quadratic dependence on window size in direct spatial convolution. Additional experiments on multiple GF-2 and Gao Fen Duo Mo scenes show that the proposed method provides stable AE correction and achieves higher image quality and visual interpretability than the compared methods in HSR imagery.

This study employs a Monte Carlo-based evaluation method to analyze the polarimetric measurement accuracy of the Polarized Scanning Atmospheric Corrector (PSAC) onboard the HJ-2A/B satellites. The conventional approach to evaluating polarization measurement accuracy is limited by its simplistic deviation analysis between the analytical results and theoretical true values, neglecting the propagation of uncertainties from multiple calibration parameters. To address this, we propose a systematic Monte Carlo-based method to comprehensively assess the polarization measurement system’s accuracy by statistically incorporating the error distributions of all calibration parameters. The Monte Carlo method leverages probability theory to simulate measurement processes through pseudo-random sampling from parameter error distributions, systematically quantifying polarization accuracy uncertainties. The analysis incorporates eight critical calibration parameters—scale factors, instrumental polarization, angular deviations, and relative gain ratios—to evaluate system performance. Monte Carlo simulations demonstrate that the PSAC achieves a polarimetric accuracy of better than 0.0035 and 0.005 for polarization states of P ⩽ 0.2 and P ⩽ 0.3, respectively. These results exhibit strong consistency with the conventional approach, thereby validating the effectiveness of the proposed method. Furthermore, the study establishes a robust framework for evaluating the polarimetric accuracy of spaceborne scanning polarimeters, which ultimately enhances the reliability of atmospheric correction and environmental monitoring applications.

Dark-field scattering inspection is one of the commonly employed methods for detecting surface defects on optical components. The amount of scattered energy collected by the inspection system is strongly affected by defect size, the incidence angle of the illumination, and the polarization state. In this study, an electromagnetic simulation model of dark-field scattering from surface defects on curved optical elements was constructed using the finite-difference time-domain (FDTD) method. A series of systematic FDTD simulations were carried out to analyze the scattered-field distributions of curved optics under varying defect sizes, illumination angles, and polarization states. The results provide practical guidance for the optimization of dark-field inspection of curved optical components.

This paper designs a polarization imaging electronics system based on the Sony IMX250MZR polarization image sensor and FPGA architecture, focusing on resolving detector readout configuration and multi-channel LVDS dynamic phase alignment issues. In the FPGA control logic design, the detector readout configuration process is clarified, and a multi-channel LVDS adjustment method based on dynamic phase alignment is optimized, achieving steady-state transmission of multi-channel data through adaptive timing compensation. A high-speed DDR2-based data caching architecture is established, and efficient interaction with peripheral devices is accomplished via the Camera Link protocol. Signal-to-noise ratio (SNR) is evaluated based on a SNR model, as well as its nonlinearity errors, enabling full dynamic range analysis of system performance in laboratory. The results indicate that the system’s SNR and linearity performance reach the level of Sony industrial cameras equipped with the same detector. The study further confirms the effectiveness of the detector readout configuration and multi-channel LVDS dynamic compensation, providing a scalable hardware architecture and quantitative evaluation framework for electronic design in such polarimetric imaging systems.