Limin Zou

We propose parallel image-scanning autocorrelation-deconvolution microscopy (PISADM), a multifocal-scanning strategy that converts global reconstruction into spatially localized reconstructions. By combining spatially localized multifocal excitation with scanning-position-wise SACD reconstruction and focus-wise pixel reassignment, PISADM reduces the effective accumulation of finite-frame cross terms and improves reconstruction stability in dense regions, thereby enhancing the separability of closely spaced structures. Monte-Carlo simulations show that, at an SBR of 10 dB, PISADM reduces the stable separability limit to approximately 100–110 nm, compared with about 140–150 nm for SACD under the same conditions. Fourier ring correlation analysis shows that PISADM preserves more reliable high-frequency information near the separability limit. In two-dimensional, long-term, and three-dimensional imaging of PCSTAR-labeled microtubules, PISADM consistently provides clearer separation of neighboring filaments while better preserving structural fidelity. These results indicate that PISADM provides a practical strategy for resolving densely packed structures with small spatial separations in super-resolution imaging.

Multimodal remote sensing data is inevitably affected by noise due to atmospheric conditions, sensor limitations, and other factors. However, existing deep learning-based multimodal remote sensing classification (MRSC) methods overlook the impact of these data noise, which produces the uncertainty and decreases classification accuracy. To address this problem, this paper first explores uncertainty-based dendritic model (UDM) for MRSC, which reduces the uncertainty at single-modality feature extraction and multimodal feature fusion stages. At the single-modality feature extraction stage, dendrites, as a novel type of neurons, have been demonstrated with strong feature extraction abilities. Inspired by the dendritic structure, we first design a dendritic-based spatial-channel feature extraction (DSCE) module. Specifically, a dendritic neural layer (DNL) is designed in DSCE. The proposed DNL constructs a multi-branch fusion strategy to enhance the expressive capacity of multimodal remote sensing feature extraction, which achieves the localized subspace computation ability. Furthermore, based on the extracted features by DSCE, a dendritic uncertainty-based feature enhancement (DUFE) module is explored to reduce the uncertainty from the extracted features. DUFE exploits uncertainty estimation to adaptively refine the extracted representations, thereby improving feature robustness and discriminative power for MRSC. At the multimodal feature fusion stage, considering the feature redundancy of the different modalities, a dendritic-based uncertainty-aware fusion (DUAF) module is proposed. DUAF performs feature fusion by dynamically assigning weights based on the estimated uncertainty of each modality, thus enhancing classification performance. Experiments on benchmark datasets demonstrate that the proposed UDM outperforms current state-of-the-art methods based on Transformer, convolutional neural network, and Mamba for MRSC. The code is available at https://github.com/hx0558/UDM.

XinMultimodal remote sensing data always comprise hyperspectral image, light detection and ranging data, and synthetic-aperture radar. Different modalities supply complementary information that improve accuracy for multimodal remote sensing classification. Although deep learning-based methods have been a mainstream, they also introduce high computational and energy cost. Unlike existing models, spiking neural network (SNN) is intrinsically energy-efficient, which reduces computation cost and energy expenditure with only a small subset of neurons. However, SNN extracts spatial and channel features without considering the redundancy among different remote sensing modalities. To reduce cross-modal redundancy while retaining important features, this paper proposes an efficient spatial-channel spiking neural network for multimodal remote sensing data classification. Firstly, in the modality fusion step, considering the quality of different modalities is various (i.e., HSI is impaired by cloud), an energy-guided multimodal remote sensing fusion strategy is proposed, which allocates the high weight for the informative single modality suppressing less-informative ones by optimizing the generation bound. Secondly, we leverage information from the high-quality modality. In the spatial feature learning step, an efficient spatial SNN is proposed. It transforms the spatial features to the frequency domain, which shares parameters across different time steps in the spatial dimension to reduce spatial feature redundancy. Finally, to further reduce the redundancy, an efficient channel SNN is explored, it focuses on learning important spike representations in the channel dimension by learning learnable parameters. Experimental results on the three multimodal remote sensing datasets indicate that the proposed methods are competitive compared to the state-of-the-art models.

The embedded ultra-precision displacement measurement is of great interest in developing high-end equipment as well as precision metrology. However, conventional interferometers only focus on measurement accuracy neglecting the sensor volume and requirement of embedded measurement, thus hindering their broad applications. Here we present a new sensing method for realizing large-range displacement measurement in narrow space scenarios based on the combination of a fiber microprobe interference-sensing model and precision phase-generated carrier. This is achieved by microprobe tilted-axis Gaussian optical field diffraction and high-order carrier demodulation to realize large-range displacement sensing. It is uncovered that the microprobe element misalignment and phase demodulation means play pivotal roles in the interference signal and the accuracy of large-range displacement sensing. The analysis shows that the proposed interference-sensing method can effectively reduce the nonlinearities. Experimental results illustrate that the measurement range extends from 0 to 700 mm. Furthermore, the maximum nonlinear error is reduced from tens of nanometers to 0.82 nm over the full range, allowing subnanometer accuracy for embedded measurements in the hundreds of millimeters range.

Aiming at the problem of small measurement range and difficult measurement of traditional contact sensor method in the rotating state of aero-engine blades, this paper proposes a novel deformation measurement method for rotating blade based on image correlation matching and mismatch correction. Firstly, a Particle Swarm Optimization-Iterative Local Search image intelligent matching algorithm is proposed, which effectively balances global search and local optimization, and the image matching displacement accuracy reaches 10⁻³ pixel. Secondly, a mismatch point detection method based on multi-scale local Root Mean Square is proposed, and the high-precision detection of mismatch points is realized by considering the influence of local structure. Finally, through the hierarchical refinement of the sub-pixel level mismatch point correction method, the mismatch point is corrected in the two iterative stages of global search and local optimization. The corrected mismatch point response value K is reduced by 99% compared with that before correction, which further improves the accuracy of deformation field calculation. In the experiment, the deformation of the rotating blade of 4500 RPM is measured, which proves that the image intelligent matching algorithm and the deformation field calculation method proposed in this paper can provide new methods and technical support for the accurate measurement of the blade surface deformation.

As process nodes of advanced integrated circuits continue to decrease below 10 nm, the requirement for overlay accuracy is becoming stricter. The alignment sensor measures the position of the alignment mark relative to the wafer; thus, sub-nanometer alignment position accuracy is vital. The Phase Grating Alignment (PGA) method is widely used due to its high precision and stability. However, the alignment error caused by the mark asymmetry is the key obstacle preventing PGA technology from achieving sub-nanometer alignment accuracy. This error can be corrected using many methods, such as process verification and multi-channel weighted methods based on multi-diffraction, multi-wavelength and multi-polarization state alignment sensors. However, the mark asymmetry is unpredictable, complex and difficult to obtain in advance. In this case, the fixed-weight method cannot effectively reduce the alignment error. Therefore, an adaptive weighted method based on the error distribution characteristic of a multi-channel is proposed. Firstly, the simulation result proves that the error distribution characteristic of the multi-alignment result has a strong correlation with the mark asymmetry. Secondly, a concrete method of constructing weight values based on error distribution is described. We assume that the relationship between the weight value of each channel and the deviations of all channels’ results is second-order linear. Finally, without other prior process correction in the simulation experiment, the residual error’s Root Mean Square (RMS) of fixed weighted method is 14.0 nm, while the RMS of the adaptive weighted method is 0.01 nm, when dealing with five typical types of mark asymmetry. The adaptive weighted method exhibits a more stable error correction effect under unpredictable and complicated mark asymmetry.

To produce a standard ultralow frequency (ULF) vibration waveform for calibration with an electromagnetic vibrator (EMV) is challenging because of its tiny electromagnetic driving force and low signal-to-noise ratio (SNR). This article presents a model-free compensation method for the micro disturbance of the EMV, which mainly originates from nonlinear forces generated by the bending or tensioning of pipes and wires. The influence of micro disturbance on the system is analyzed to extract the disturbance current without estimating the equivalent model and compensated in the input; thus, the force sensorless and model-free compensation of micro disturbance is realized in real-time. Experimental results in the frequency range of 0.001–1 Hz indicate that the proposed method reduces the displacement tracking error by 10.14% and the total harmonic distortion (THD) by 19.44% relative to the model-based compensation method. The proposed method effectively reduces the THD of the ULF vibration waveform with a low computational burden. It can be widely used for ULF vibration generation in engineering applications.