Search

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 multichannel separated cross-attention (MSCA) and full-feature joint cross-attention (FJCA) 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 (ATDWM) 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 are available at https://github.com/LiupengLin/CAFE. [ABSTRACT FROM AUTHOR]

• A fusion model is proposed to sharpen Sentinel-2 20/60-m bands to 10-m. • A hierarchical fusion framework is applied to maximize the use of observed information. • Experiments demonstrate the remarkable sharpening performance of the model. • The created Sentinel-2 all-band 10-m data have the potential for a variety of land services. Earth observations from the Sentinel-2 mission have been extensively accepted in a variety of land services. The thirteen spectral bands of Sentinel-2, however, are collected at three spatial resolutions of 10/20/60 m, and such a difference brings difficulties to analyze multispectral imagery at a uniform resolution. To address this problem, we developed a hierarchical fusion network (HFN) to sharpen 20/60-m bands and generate Sentinel-2 all-band 10-m data. The deep learning architecture is used to learn the complex mapping between multi-resolution input and output data. Given the deficiency of previous studies in which the spatial information is inferred only from the fine-resolution bands, the proposed hierarchical fusion framework simultaneously leverages the self-similarity information from coarse-resolution bands and the spatial structure information from fine-resolution bands, to enhance the sharpening performance. Technically, the coarse-resolution bands are super-resolved by exploiting the information from themselves and then sharpened by fusing with the fine-resolution bands. Both 20-m and 60-m bands can be sharpened via the developed approach. Experimental results regarding visual comparison and quantitative assessment demonstrate that HFN outperforms the other benchmarking models, including pan-sharpening-based, model-based, geostatistical-based, and other deep-learning-based approaches, showing remarkable performance in reproducing explicit spatial details and maintaining original spectral features. Moreover, the developed model works more effectively than the other models over the heterogeneous landscape, which is usually considered a challenging application scenario. To sum up, the fusion model can sharpen Sentinel-2 20/60-m bands, and the created all-band 10-m data allows image analysis and geoscience applications to be authentically carried out at the 10-m resolution. [ABSTRACT FROM AUTHOR]

Cloud contamination greatly limits the potential utilization of optical remote sensing images for geoscience applications. Many solutions have been developed to remove the clouds from multispectral images. Among these approaches, the temporal-based methods which borrow complementary information from multitemporal images outperform the other methods. However, the common fundamental supposition of the temporal-based methods decides that they are only suitable for scenes with phenological changes, while they perform poorly in cases with significant land cover changes. In this paper, a cloud removal procedure based on multisource data fusion is developed to overcome this limitation. On the basis of the temporal-based approaches, which employ a cloud-free image as reference, this method further introduces two auxiliary images with similar wavelengths and close acquisition dates to the reference and target (contaminated) images into the reconstruction process. The temporal variability of the land cover is captured from the two auxiliary images through a modified spatiotemporal data fusion model, and thus, the serious errors produced by the temporal-based methods can be avoided. Moreover, a residual correction strategy based on the Poisson equation is used to enhance the spectral coherence between the recovered and remaining regions. The experiments confirmed that the proposed method can perform very well for cases with significant land cover changes. Compared with some state-of-the-art approaches, it produces lower bias and more robust efficacy. In conclusion, our method will act as an important technical supplement to the current cloud removal framework, and it provides the possibility to handle scenes with significant land cover changes. [ABSTRACT FROM AUTHOR]