Kunlun Qi

The fusion of optical and SAR remote sensing imagery has become increasingly crucial for accurate multi-label remote sensing scene classification (MRSSC), which plays an essential role in producing reliable land use and land cover (LULC) products. However, visual heterogeneity between optical and SAR data, together with the speckle noise inherent in SAR imagery, greatly limits the performance of existing multimodal fusion approaches. To overcome these challenges, this paper proposes a Frequency-Domain Progressive Fusion Network (FDPFNet) that adopts a hybrid CNN–Transformer architecture to serve as an effective and unified multimodal backbone for MRSSC. First, a Low-Frequency Convolution (LFConv) block is introduced, utilizing wavelet transform to highlight low-frequency components shared across modalities while suppressing high-frequency noise in SAR data. Second, a Two-Frequency Decomposition (TFD) block is designed to decompose features into high- and low-frequency components, allowing comprehensive fusion of modality-shared low-frequency semantics while mitigating the adverse effects of inconsistent high-frequency details. Finally, an Adaptive Feature Fusion (AFF) block is developed to dynamically balance intra-modal feature consistency and inter-modal complementarity across multiple hierarchical levels, thereby achieving more effective optical–SAR fusion. Extensive experiments conducted on the BigEarthNet-MM and SEN12-MLRS datasets demonstrate that FDPFNet consistently outperforms state-of-the-art methods, and the ablation studies further verify the effectiveness of each proposed module and the overall architecture.

Remote sensing imagery is a fundamental data source in spatial information science and is widely used in earth observation and geospatial applications. The explosive growth of such data poses significant challenges for online management and service, particularly in terms of storage scalability, processing efficiency, and real-time accessibility. To overcome these limitations, we propose DDMS, a distributed data management and service framework for heterogeneous remote sensing data that structures its functionality around three core components: storage, computing, and service. In this framework, a distributed integrated storage model is constructed by integrating file systems with database technologies to support heterogeneous data management, and a parallel computing model is designed to optimize large-scale image processing. To verify the effectiveness of the proposed framework, a prototype system was implemented and evaluated with experiments on representative datasets, covering both optical and InSAR images. Results show that DDMS can flexibly adapt to heterogeneous remote sensing data and storage backends while maintaining efficient data management and stable service performance. Stress tests further confirm its scalability and consistent responsiveness under varying workloads. DDMS provides a practical and extensible solution for large-scale online management and real-time service of remote sensing images. By enhancing modularity, scalability, and service responsiveness, the framework supports both research and practical applications that depend on massive earth observation data.

Steel surface defect detection is a crucial step in ensuring industrial production quality. However, due to significant variations in scale and irregular geometric morphology of steel surface defects, existing detection algorithms show notable deficiencies in multi-scale feature representation and cross-layer multi-scale feature fusion efficiency. To address these challenges, this paper proposes an improved real-time steel surface defect detection model, NGD-YOLO, based on YOLOv5s, which achieves fast and high-precision defect detection under relatively low hardware conditions. Firstly, a lightweight and efficient Normalization-based Attention Module (NAM) is integrated into the C3 module to construct the C3NAM, enhancing multi-scale feature representation capabilities. Secondly, an efficient Gather–Distribute (GD) mechanism is introduced into the feature fusion component to build the GD-NAM network, thereby effectively reducing information loss during cross-layer multi-scale information fusion and adding a small target detection layer to enhance the detection performance of small defects. Finally, to mitigate the parameter increase caused by the GD-NAM network, a lightweight convolution module, DCConv, that integrates Efficient Channel Attention (ECA), is proposed and combined with the C3 module to construct the lightweight C3DC module. This approach improves detection speed and accuracy while reducing model parameters. Experimental results on the public NEU-DET dataset show that the proposed NGD-YOLO model achieves a detection accuracy of 79.2%, representing a 4.6% mAP improvement over the baseline YOLOv5s network with less than a quarter increase in parameters, and reaches 108.6 FPS, meeting the real-time monitoring requirements in industrial production environments.

The urgency of global climate change has driven the rapid expansion of photovoltaic (PV) energy systems. However, accurately identifying PV panels remains a major challenge due to complex environmental backgrounds, spectral confusion, and the lack of high-quality annotated datasets. These factors significantly impact the generalization ability of deep learning models in large-scale high-resolution remote sensing applications, thereby limiting the effective monitoring and planning of PV power stations. To address these challenges, this article proposes a highly adaptable PV panel segmentation network, detailed PV monitoring (DPVM), specifically designed to enhance PV panel recognition in high-resolution imagery. DPVM integrates an adaptive context-aware module (ACAM) and a deep feature reconstruction decoder (DFRD). ACAM improves segmentation accuracy by leveraging multiscale feature fusion and spatial attention mechanisms. DFRD employs multistage decoding and feature synthesis to achieve high-quality image reconstruction. We trained DPVM on our self-constructed Northwest China PV dataset to ensure comprehensive learning of PV panel characteristics. Subsequently, we conducted generalization tests on other publicly available datasets, including AIR-PV and PVP. Experimental results demonstrate that DPVM exhibits outstanding robustness and broad adaptability, ensuring stable performance across diverse scenarios. Specifically, DPVM excels in complex backgrounds, significantly reducing PV panel missed detections, improving edge delineation, and outperforming classical and state-of-the-art segmentation models in key metrics.

Scene classification is an active research area in the remote sensing (RS) domain. Some categories of RS scenes, such as medium residential and dense residential scenes, would contain the same type of geographical objects but have various spatial distributions among these objects. The adjacency and disjointness relationships among geographical objects are normally neglected by existing RS scene classification methods using convolutional neural networks (CNNs). In this study, a multi-output network (MopNet) combining a graph neural network (GNN) and a CNN is proposed for RS scene classification with a joint loss. In a candidate RS image for scene classification, superpixel regions are constructed through image segmentation and are represented as graph nodes, while graph edges between nodes are created according to the spatial adjacency among corresponding superpixel regions. A training strategy of a jointly learning CNN and GNN is adopted in the MopNet. Through the message propagation mechanism of MopNet, spatial and topological relationships imbedded in the edges of graphs are employed. The parameters of the CNN and GNN in MopNet are updated simultaneously with the guidance of a joint loss via the backpropagation mechanism. Experimental results on the OPTIMAL-31 and aerial image dataset (AID) datasets show that the proposed MopNet combining a graph convolutional network (GCN) or graph attention network (GAT) and ResNet50 achieves state-of-the-art accuracy. The overall accuracy obtained on OPTIMAL-31 is 96.06% and those on AID are 95.53% and 97.11% under training ratios of 20% and 50%, respectively. Spatial and topological relationships imbedded in RS images are helpful for improving the performance of scene classification.

Ecological environment quality is a long-term continuous concept that is affected by various environmental factors. Its assessment has important implications for implementing the planning and protection of dynamic regional ecosystems. Therefore, this study attempted to obtain these indicators (green, dry, wet, heat) through the Google Earth Engine (GEE) platform, and then coupled the ecological environment quality index in the middle reaches of the Yangtze River Basin (MYRB) between 2000 and 2019, based on the remote sensing ecological index (RSEI). The major results show that: (1) changes in the four indicators in summer were more obvious than those in winter, and the changes were concentrated in the central and northern regions of the MYRB; (2) both the modified normalized difference water index (MNDWI) and normalized differential build-up and bare soil index (NDBI) in summer and winter have higher weighting ratios, implying that water body changes and human activities had a greater impact on the ecological environment; and (3) ecological environment quality in the MYRB between 2000 and 2019 was relatively flat. The ecological conditions began to deteriorate in 2008, and substantial ecological degradation was noted in some areas between 2008 and 2019 (18.7% in the central region, 16.0% in the eastern region). The MYRB has an important position in the Yangtze River economic belt and is an important part of the Yangtze River protection. This research could provide a theoretical basis and decision support for the development and protection of the Yangtze River Basin (YRB) green economy.

Remote sensing imagery typically comprises successive background contexts and complex objects. Global average pooling is a popular choice to connect the convolutional and fully connected (FC) layers for the deep convolution network. This article equips the networks with another pooling strategy, namely the deep object-centric pooling (DOCP), to pool convolutional features considering the location of an object within the scene image. The proposed DOCP network structure consists of the following two steps: inferring object's location and separately pooling the foreground and background features to generate an object-level representation. Specifically, a spatial context module is presented to learn the location of the object of interest in the scene image. Then, the convolutional feature maps are pooled separately in the foreground and background of the object. Finally, the FC layer concatenates these pooled features and is followed by a batch normalization layer, a dropout layer, and a softmax layer. Two challenging datasets are employed to validate our approach. The experimental results demonstrate that the proposed DOCP-net can outperform the corresponding pooling methods and achieve a better classification performance than other pretrained convolutional neural network-based scene classification methods.