Roads are closely related to people’s lives, and road network extraction has become one of the most important remote sensing tasks. This study aimed to propose a road extraction network with an embedded attention mechanism to solve the problem of automatic extraction of road networks from a large number of remote sensing images. Channel attention mechanism and spatial attention mechanism were introduced to enhance the use of spectral information and spatial information based on the U-Net framework. Moreover, residual densely connected blocks were introduced to enhance feature reuse and information flow transfer, and a residual dilated convolution module was introduced to extract road network information at different scales. The experimental results showed that the method proposed in this study outperformed the compared algorithms in overall accuracy. This method had fewer false detections, and the extracted roads were closer to ground truth. Ablation experiments showed that the proposed modules could effectively improve road extraction accuracy.

The data fusion technology aims to aggregate the characteristics of different data and to obtain products with multiple data advantages. To solve the problem of reduced resolution of polarimetric synthetic aperture radar (PolSAR) images due to system limitations, we propose a fully PolSAR images and single-polarization synthetic aperture radar (SinSAR) images fusion network to generate high-resolution PolSAR (HR-PolSAR) images. To take advantage of the polarimetric information of the low-resolution PolSAR (LR-PolSAR) images and the spatial information of the high-resolution single-polarization SAR (HR-SinSAR) images, we propose a fusion framework for joint LR-PolSAR images and HR-SinSAR images and design a cross-attention mechanism to extract features from the joint input data. Besides, based on the physical imaging mechanism, we designed the PolSAR polarimetric loss functions for constrained network training. The experimental results confirm the superiority of the fusion network over traditional algorithms. The average peak signal-to-noise ratio (PSNR) is increased by more than 3.6 dB, and the average mean absolute error (MAE) is reduced to less than 0.07. Experiments on polarimetric decomposition and polarimetric signature show that it maintains polarimetric information well.

Under foggy conditions, the atmospheric scattering effect diminishes the illumination intensity in images, causing a decline in the contrast of remote sensing images, which impacts the accuracy of object detection algorithms. Mainstream algorithms typically employ image restoration methods as a preprocessing step before object detection, however, the distributed method increases the time taken by the model to perform inference and it is difficult to ensure the dehazing module has positive effects on the object detection network modeling. To this end, we propose a Multi-Scale feature collaborative Foggy image Object Detection Network (MS-FODN) within a multi-task learning framework. This end-to-end model integrates a dehazing module to assist in object detection. In the proposed network, the encoder parameters of both the dehazing and detection modules are shared during the encoding stage. For the decoding stage, the dehazing module employs the UNet architecture with a multi-constrained loss function to enhance dehazing, while the detection module adopted the DiffusionDet decoder as its detection head. To address the challenge of substantial scale differences in remote sensing targets, we proposed a multi-scale attention feature fusion module that refines and calibrates features at different scales and aggregates them through the feature pyramid module. Additionally, we implemented a multi-scale KL loss optimization to minimize the negative impacts on object detection outcomes caused by the suboptimal performance of the dehazing decoder. Given the absence of open-source foggy aerial datasets, we developed a real remote sensing object detection dataset, RH-Vehicle, captured by UAVs. MS-FODN has been proven to exceed recent state-of-the-art methods in detection accuracy through extensive experiments on both simulated and real datasets.

Dense medium-resolution images play an important role in time-series geoscience applications. However, due to technical limitations, remote sensing imaging systems inevitably trade off temporal frequency and spatial swaths, resulting in difficulties to acquire images simultaneously with high spatial and temporal resolution. To overcome this limitation, under the framework of residual learning, we propose a Cross-attention based Adaptive weighting Fusion nEtwork (CAFE) for MODIS-Landsat spatiotemporal fusion to generate dense medium-resolution images. Based on the cross-attention mechanism, we propose multi-channel separated cross-attention and full-feature joint cross-attention blocks to enhance spatial resolution and retain spectral signatures from the perspectives of band-wise processing and full-feature joint processing, respectively. The adaptive temporal difference weighting mechanism is proposed to improve the ability to capture dynamic land surface changes. Besides, we employ an adaptive fusion loss function to constrain the network training. Experimental results indicate that the developed method is superior to several existing algorithms in terms of visual evaluation and quantitative evaluation and it can generate high-quality fusion results in scenarios of both subtle and dramatic temporal changes. Codes will be available at https://github.com/LiupengLin/CAFE.