Search Results (814)

Cloud detection is essential for quantitative land-surface remote sensing and cloud-climate research. However, existing methods often prioritize spatial features over spectral features, which limits thin-cloud detection. To address this issue, this paper proposes a Thin-Cloud-Sensitive Network (TCSNet) for hyperspectral imagery. TCSNet employs an encoder–decoder architecture with a dual-branch design: a convolutional neural network (CNN) extracts multi-scale local features, while a PVTv2-B2 Transformer captures long-range spectral dependencies. To effectively integrate the complementary representations from both branches, a Cross-Modal Fusion (CMF) module with a lightweight single-channel gate is introduced at each stage, followed by a channel attention mechanism (SE) for feature recalibration. Subsequently, a Multi-Scale Fusion (MSF) module is used to integrate multi-level features through a top-down pathway, enabling deep semantic information to guide shallow feature expression. Furthermore, to enhance the decoder’s feature representation capability, a Combined Attention Mechanism (CAM) is incorporated at each decoder stage. This design enables the network to simultaneously focus on important channels, salient regions, and cloud boundaries, effectively alleviating spectral confusion between thin clouds and the underlying surface. Experimental results on Gaofen-5 01 hyperspectral data demonstrate that TCSNet achieves the highest recall (92.98%), ${\mathrm{R}\mathrm{e}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{l}}_{thin}$ (85.59%), and ${\mathrm{R}\mathrm{e}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{l}}_{thick}$ (99.75%), thereby validating its superiority for thin-cloud detection. Full article

The unchecked expansion of tea plantations onto steep, forest-adjacent slopes in subtropical mountains engenders a conflict between agricultural productivity and ecosystem integrity, particularly by exacerbating habitat fragmentation and soil erosion. While precise monitoring is essential to navigate this trade-off for sustainable management, accurate inventorying remains a challenge due to the plantations’ strong phenological variability, heterogeneous canopy structures, and high spectral confusion with surrounding vegetation. This study proposes a feature-optimized deep learning framework for mapping and characterizing tea plantations in complex landscapes, using Xinyang City, China, as a study area. The framework integrates multi-temporal Sentinel-1/2 observations with a sequential Jeffries-Matusita (JM)-Pearson feature filtering strategy. This approach effectively condenses a 132-variable high-dimensional pool (including optical spectra, vegetation indices, textures, and SAR polarimetry) into a compact 28-feature subset (a 78.8% reduction), preserving critical phenological and structural cues while minimizing redundancy. These optimized predictors drive a hybrid VGG16–UNet++ segmentation network, which couples transfer-learning-based semantic encoding with detail-preserving dense skip fusion. Extensive experiments across 18 model–feature configurations demonstrate that the optimal setting achieves an Overall Accuracy of 97.82%, an F1-score of 0.9093, and a mean IoU of 0.7968. Notably, the method significantly reduces misclassification in rugged, cloud-prone terrain, yielding a User’s Accuracy of 91.14% for tea. Based on the generated wall-to-wall map, we derived two decision-support indicators: multi-threshold steep-slope exposure and a normalized tea–forest interface density. This framework provides actionable, high-precision spatial products to support slope-based zoning, ecological restoration, and sustainable management in fragile mountain agroforestry systems. Full article

Semantic segmentation of land use and land cover (LULC) in arid regions remains challenging due to severe class imbalance, fragmented spatial distributions, and high spectral similarity among different land cover types. These characteristics often lead to an information bottleneck in deep segmentation networks and hinder the extraction of discriminative semantic representations. To address these issues, we propose SDS-Former, a lightweight semantic segmentation network specifically designed for remote sensing imagery in arid environments. SDS-Former incorporates an SSM-inspired Lightweight Semantic Enhancement (LSE) module to strengthen contextual modeling and alleviate the loss of discriminative information in deep features. To tackle scale variations, a Dynamic Selective Feature Fusion (DSFF) module is employed in the decoder to adaptively weight and fuse high-level semantics with low-level spatial details. Furthermore, a Feature Refinement Head (FRH) is introduced to enhance boundary localization and improve the recognition of small-scale and sparsely distributed land cover objects. Extensive ablation and comparative experiments demonstrate that SDS-Former consistently outperforms representative semantic segmentation methods across multiple evaluation metrics. On the Tarim Basin dataset, the proposed network achieves a mean Intersection over Union (mIoU) of 82.51% and an F1 score of 86.47%, indicating its superior effectiveness and robustness. Qualitative results further verify that SDS-Former exhibits clear advantages in distinguishing spectrally similar land cover types and preserving the spatial continuity of ground objects in complex arid-region scenes. Full article

With the continuous advancement of polarization spectral sensing technology, multi-band polarization image fusion has emerged as a novel approach to image fusion. By integrating spectral and polarization information, this method overcomes the limitations of relying on a single information source and significantly improves overall image quality. To address this, this paper proposes a new polarization spectral fusion algorithm. First, feature matching is employed to achieve pixel-level spatial alignment of multi-band polarization images. Then, a fusion strategy based on multi-scale decomposition and singular value decomposition is adopted to preserve structural information and fine details. Subsequently, fractional-order processing and guided filtering are applied to enhance details and suppress noise. Finally, a progressive reconstruction from low to high scales is performed to ensure hierarchical consistency and information integrity throughout the fusion process. In addition, spectral information is utilized for color restoration, enabling the final image to achieve high spatial resolution while maintaining natural and rich color representation.Experimental results demonstrate that the proposed method effectively integrates features from different spectral bands and polarization information while preserving maximum similarity, leading to significant improvements in both image quality and detail representation. Full article

Haul roads in surface mining are critical infrastructure directly influencing operational productivity, safety, and costs. However, these networks change frequently due to ongoing mining activities, making traditional mapping methods impractical for large-scale or rapidly evolving sites. Remote sensing imagery offers a scalable alternative, yet complex backgrounds, variable road widths, and spectral similarities between roads and surrounding surfaces make accurate extraction challenging. This study proposes HRM-Net, a hybrid transformer–CNN autoencoder framework for automated extraction of mine haul roads from remote sensing imagery. HRM-Net introduces inception-like patch embedding to capture local contextual information and employs a manifold-constrained hyper-connection strategy in the attention and fusion blocks to enhance information flow across the architecture. This hierarchical design enables progressive learning of discriminative semantic representations across multiple spatial resolutions, critical for road extraction in cluttered mining environments. Trained and evaluated on diverse mine sites, HRM-Net achieved 92.53% overall accuracy, 85.12% F1-score, 75.57% mIoU, 83.57% precision, and 86.94% recall, outperforming state-of-the-art transformer-based and CNN-based segmentation models. Furthermore, model interpretability was analyzed through linear probing and boundary alignment evaluations. Results demonstrate that discriminative features emerge at early network stages and are effectively preserved throughout the architecture, while boundary predictions exhibit superior consistency compared to existing approaches. Full article

Soil salinization is one of the primary forms of land degradation in arid and semi-arid regions, severely constraining agricultural production in Xinjiang’s oases. Unmanned aerial vehicle (UAV) imagery provides an effective means for precise monitoring of soil salinization, with image spatial resolution being a key factor affecting assessment accuracy. However, traditional single-scale remote sensing monitoring methods rely solely on spectral and textural features at the leaf scale (0.1 m resolution captures leaf-scale characteristics), neglecting the contribution of multi-scale features (single-row canopy scale and single-membrane-covered area scale (6-row crop canopy)) to soil salinity. For instance, 0.5–1 m reflects single-row canopy scale, while 2 m reflects single-membrane-covered area scale. Therefore, this study developed a multi-scale UAV imagery and machine learning framework to enhance soil electrical conductivity prediction accuracy. This study focuses on oasis cotton fields in Shaya County, Xinjiang. Based on UAV multispectral imagery, we resampled data to generate eight datasets at different spatial resolutions: 0.1, 0.5, 1, 1.5, 2, 2.5, 5, and 10 m. For each resolution, we calculated 21 spectral indices and 48 texture features to construct a feature set. At both single and multispatial scales, spectral indices, texture features, and their spectral-texture fusion features were constructed. Combining these with Backpropagation Neural Network (BPNN), Random Forest Regression (RFR), and Extreme Gradient Boosting (XGBoost) models, a soil EC estimation framework was developed. The impact of three feature combination schemes on cotton field soil conductivity estimation using single-scale UAV imagery was compared. The accuracy of soil EC estimation for cotton fields was compared between multi-spatial scale and single-scale UAV image features. The optimal combination strategy for a multi-spatial scale and multiple features was determined. Results indicate that combining spectral and texture features yields the highest estimation accuracy for cotton field soil electrical conductivity in single-scale analysis. Multi-spatial scale image features outperform single-scale image features in estimating cotton field soil electrical conductivity accuracy. By comparing different feature combinations, when integrating 0.5 m spatial-scale spectra (S1, EVI, DVI, NDVI, Int1, SI) with 0.1 m texture features (RE1_ent, R_cor, RE1_cor, G_hom, B_mea, R_con, NIR_con), the XGBoost model achieved the optimal prediction accuracy ($R^2$ = 0.693, RMSE = 0.515 dS/m), outperforming the methods using multiple features at a single scale. This study developed a novel multi-scale image feature fusion technique to construct a machine learning model. This method describes the image characteristics of soil electrical conductivity at different geographical scales, providing a reference approach for the rapid and accurate prediction of soil electrical conductivity in arid regions. Full article

Due to the inherent limitations of both hyperspectral and multispectral imagery, balancing high spatial resolution with high spectral fidelity has become one of the fundamental challenges in remote sensing image processing. A prevailing strategy is to fuse these two types of data to reconstruct images that jointly preserve their respective advantages. However, existing reconstruction approaches still suffer from complex coupling between spatial and spectral information, and limited feature extraction capabilities. To address these issues, this study proposes PMSwinNet (Pyramid Multi-scale Swin Transformer Network), a novel architecture that integrates pyramid-based feature enhancement with Transformer mechanisms. The PMSwinNet incorporates multi-scale pyramid feature fusion and window-based self-attention. Through a progressive multi-stage design and three complementary components—feature extraction and reconstruction modules—the Transformer branch leverages window partitioning and shifting operations to capture long-range spatial dependencies and local contextual cues, while the pyramid features extract both global and local information across multiple spatial scales. In addition, a high-frequency branch is introduced, which employs lightweight convolutions to enhance edges, textures, and other high-frequency details, effectively suppressing blurring and artifacts during reconstruction. Experimental evaluations on multiple public hyperspectral datasets demonstrate that the PMSwinNet outperforms state-of-the-art methods, particularly in terms of detail preservation, spectral distortion suppression, and robustness. Full article

Synthetic aperture radar (SAR) ship detection in complex inshore scenarios has long been constrained by the coupled effects of speckle noise and small–scale weak scattering targets. Although feature–level frequency–domain denoising methods partially alleviate noise interference, existing studies predominantly focus on spatial frequency modeling and implicitly assume consistent spectral responses and discriminative contributions across channels. This assumption may lead to over–suppression of weak ship targets under complex backgrounds. To address the incomplete dimensionality of current frequency–domain modeling, this paper proposes FSMD–Net, a joint spatial–channel spectral modeling framework for SAR ship detection. During multi–scale feature fusion, a coordinated modulation mechanism integrating multi–spectral channel attention with spatial frequency–domain denoising is introduced. This design enables channel discriminability and frequency–subspace denoising to act synergistically, enforcing structurally consistent spectral constraints throughout multi–scale feature propagation. Extensive experiments on SARDet–100K, HRSID, and AIR–SARShip–2.0 demonstrate that FSMD–Net achieves consistent performance improvements, particularly in small–target and strong–clutter scenarios, exhibiting enhanced detection accuracy and robustness. Full article

Land use change detection (LUCD) serves as a crucial technical cornerstone for natural resource management and ecological environment monitoring, playing an indispensable role in advancing the modernization of national governance capacities. Nonetheless, severe interference from radiometric variations on feature representation readily induces spurious changes and thus a high false alarm rate. Additionally, the challenge of balancing discriminative feature extraction and fine-grained contextual modeling leads to fragmented change regions and missed detection. To address these issues and eliminate the reliance on annotated samples, a novel framework is proposed for unsupervised LUCD, integrating superpixel-based deep feature analysis with a dense conditional random field (CRF). Firstly, relative radiometric correction and band-wise maximum stacking fusion are performed on the bi-temporal images. A simple non-iterative clustering (SNIC) algorithm is adopted to generate homogeneous superpixels with cross-temporal consistency. Then, a deep feature coupling mining mechanism is introduced to implement spatial–spectral feature extraction and in-depth parsing of invariant semantic information. Meanwhile, the difference confidence map based on dual features is constructed using superpixel-level discriminant vectors to enhance the separability. Finally, leveraging homogeneous units with spatial correspondence, a task-specific redesign of a global optimization model is established to achieve the precise extraction of change regions, which incorporates difference confidence, spatial adjacency relationship, and cross-temporal feature similarity into the dense CRF. The experimental results demonstrate that the proposed method achieves an average overall accuracy of over 90% across all datasets with excellent comprehensive performance, striking a well-balanced trade-off in practical applicability. It can effectively suppress salt-and-pepper noise, significantly improve the recall rate of change regions (maintaining at approximately 90%), and exhibit favorable superiority and robustness in complex land cover scenarios. Full article

Achieving a high frame rate and high dynamic range (HDR) under complex illumination remains a significant challenge for airborne push-broom visible-near-infrared (VNIR) hyperspectral cameras. Problematic scenarios typically include high-contrast scenes, such as ocean whitecaps alongside deep water or concurrently sunlit and shadowed urban surfaces. To address this, a real-time HDR acquisition system based on a dual-gain complementary metal–oxide–semiconductor (CMOS) image sensor is proposed. Specifically, a four-pixel HDR fusion method is developed, utilizing an optical calibration setup to accurately determine the fusion parameters and configure the spectral region of interest (ROI) for reduced data volume. The complete workflow, encompassing spectral–spatial four-pixel binning and piecewise dual-gain fusion, is implemented on a field-programmable gate array (FPGA) using a dual-port RAM-based buffering strategy and a low-latency five-stage pipeline. Experimental results demonstrate a minimal processing latency of 0.0183 ms and a maximum frame rate of 290 frames/s. By extending the output bit depth from 11 to 15 bits, the system achieves a digital dynamic range of the final output of 2.03 × 10⁴:1, representing a 9.58-fold improvement over the original low-gain data. The fused HDR data maintain high linearity and good spectral fidelity, with spectral angle mapper (SAM) values at the 10⁻³ level. Featuring a compact and low-power design, this system provides a practical engineering solution for efficient airborne VNIR hyperspectral acquisition. Full article

Synthetic Aperture Radar (SAR) and optical image fusion technology plays a crucial role in remote sensing applications. It effectively combines the high spatial resolution and rich spectral information of optical images with the all-weather and penetrating observation advantages of SAR images, thereby significantly enhancing image interpretation accuracy and task execution capabilities. This paper systematically reviews deep learning-based fusion methods for SAR and optical images, with a particular focus on recent advances in deep learning models. Furthermore, it summarizes commonly used evaluation metrics for assessing fusion image quality, providing a basis for comparing and analyzing the performance of different methods. In addition, commonly used SAR-optical fusion datasets are briefly reviewed to highlight their roles in algorithm development and performance evaluation. Unlike conventional review articles, this paper further analyzes the guidance and supporting role of fusion algorithms from the perspective of typical and specific applications. Finally, it identifies key challenges and issues faced by current fusion methods, including data registration, model lightweight design, and multimodal feature alignment, and offers perspectives on future research directions. This review aims to provide routes and references for the development of SAR and optical image fusion technology. Full article

Background/Objectives: Brain health monitoring is increasingly essential as modern cognitive load, stress, and lifestyle pressures contribute to widespread neural instability. The paper presents BrainTwin, a next-generation cognitive digital twin, as a patient-specific, constantly updating computer model that combines state-of-the-art MRI analytics for neuro-oncological assessment related to clinical study and management of tumors affecting the central nervous system (including their detection, progression, and monitoring) with real-time EEG-based brain health intelligence. Methods: Structural analysis is driven by an Enhanced Vision Transformer (ViT++), which improves spatial representation and boundary localization, achieving more accurate tumor prediction than conventional models. The extracted tumor volume forms the baseline for short-horizon tumor progression modeling. Parallel to MRI analysis, continuous EEG signals are captured through an in-house wearable skullcap, preprocessed using Edge AI on a Hailo Toolkit-enabled Raspberry Pi 5 for low-latency denoising and secure cloud transmission. Pre-processed EEG packets are authenticated at the fog layer, ensuring secure and reliable cloud transfer, enabling significant load reduction in the edge and cloud nodes. In the digital twin, EEG characteristics offer real-time functional monitoring through dynamic brainwave analysis, while a BiLSTM classifier distinguishes relaxed, stress, and fatigue states, which are probabilistically inferred cognitive conditions derived from EEG spectral patterns. Unlike static MRI imaging, EEG provides real-time brain health monitoring. The BrainTwin performs EEG–MRI fusion, correlating functional EEG metrics with ViT++ structural embeddings to produce a single risk score that can be interpreted by clinicians to determine brain vulnerability to future diseases. Explainable artificial intelligence (XAI) provides clinical interpretability through gradient-weighted class activation mapping (Grad-CAM) heatmaps, which are used to interpret ViT++ decisions and are visualized on a 3D interactive brain model to allow more in-depth inspection of spatial details. Results: The evaluation metrics demonstrate a BiLSTM macro-F1 of 0.94 (Precision/Recall/F1: Relaxed 0.96, Stress 0.93, Fatigue 0.92) and a ViT++ MRI accuracy of 96%, outperforming baseline architectures. Conclusions: These results demonstrate BrainTwin’s reliability, interpretability, and clinical utility as an integrated digital companion for tumor assessment and real-time functional brain monitoring. Full article

Accurate cloud detection remains a significant challenge due to the spectral ambiguity between clouds and bright or heterogeneous surfaces (e.g., snow, desert). While multi-angle and polarization data offer rich information, the discriminative power of joint spectral analysis for resolving these ambiguities has been underexploited. In this work, we demonstrate that physically motivated spectral band ratios and differences can robustly enhance cloud signatures. Motivated by this insight, we propose a novel deep learning framework, the Multi-angle Polarization Feature Pyramid Structure (MP-FPS), that explicitly leverages joint spectral features as discriminative priors. Our architecture employs a dual-branch network to disentangle and adaptively fuse spectral and multi-angle polarization modalities. Within this framework, a hierarchical, multi-scale cross-channel multi-angle fusion module dynamically captures spatial–spectral–angular dependencies, enriching the structural representation of clouds. Furthermore, a channel-space dual-path attention mechanism refines sub-pixel responses, significantly improving detection accuracy in challenging regions such as cloud edges and thin cirrus. Evaluated on the global POLDER-3 dataset, MP-FPS achieves a mean Intersection over Union (mIoU) of 0.8662 across diverse surface types, surpassing the official baseline by 12.4%. This study establishes joint spectral analysis as a critical enabler for high-precision cloud masking, and demonstrates its synergistic value when integrated with multi-angle polarimetric information in a unified deep architecture. Full article

Cross-domain few-shot hyperspectral image (HSI) classification is challenging due to limited labeled samples and distribution shifts across sensors and acquisition scenes, which often degrade feature representation and classification performance. This study proposes a lightweight hierarchical CNN–Transformer framework, termed LGDAF-Net (Lightweight Global and Local Dual Attention Fusion Network), for effective cross-domain few-shot HSI classification. The framework progressively enhances spectral–spatial representation through three stages: spectral–spatial feature recalibration, local spatial structure perception, and global contextual modeling. Specifically, a spectral–spatial dual-attention enhancement module (SESA) is introduced to emphasize informative spectral responses and suppress redundancy. A Local Attention Spatial Perception Module (LASPM) is designed to capture fine-grained spatial structures, while a lightweight Transformer-based Global Attention Context Modeling Module (GACM) models long-range spatial dependencies. In addition, kernel triplet loss and domain adversarial learning are incorporated to improve feature discrimination and promote cross-domain feature alignment. Experimental results on three benchmark datasets demonstrate that the proposed method achieves competitive performance compared with existing methods. Full article

Hyperspectral image (HSI) classification requires the extraction of discriminative features from high-dimensional spatial–spectral data. While the Mamba architecture has shown promise in long-sequence modeling with linear complexity, its application to HSI remains constrained by two major hurdles: the unidirectional causal scanning which fails to capture non-causal global dependencies, and the serialization-induced loss of two-dimensional spatial topology and local textures. To overcome these limitations, we propose HAMamba, a novel Hybrid Attention State Space Model. HAMamba facilitates deep representation learning through two core components: a Multi-Scale Dynamic Fusion (MSDF) module and a Hybrid Attention Mamba Encoder (HAME). Specifically, the MSDF module augments spatial perception through parallelized feature extraction and dynamically weighted integration. The HAME synergizes a Bidirectional Sequence Scan Mamba (BSSM) to establish global semantic context and a Spatial–Spectral Gated Attention (SSGA) module to refine local structural details. Comprehensive experiments on four public benchmark datasets demonstrate that the proposed HAMamba significantly outperforms state-of-the-art approaches, achieving a superior balance between classification accuracy and computational efficiency. Full article