Search

Cloud contamination is a common issue that severely reduces the quality of optical satellite images in remote sensing fields. With the rapid development of deep learning technology, cloud contamination is expected to be addressed. In this paper, we propose Denoising Diffusion Probabilistic Model-Cloud Removal (DDPM-CR), a novel cloud removal network that can effectively remove both thin and thick clouds in optical image scenes. Our network leverages the denoising diffusion probabilistic model (DDPM) architecture to integrate both clouded optical and auxiliary SAR images as input to extract DDPM features, providing significant information for missing information retrieval. Additionally, we propose a cloud removal head adopting the DDPM features with an attention mechanism at multiple scales to remove clouds. To achieve better network performance, we propose a cloud-oriented loss that considers both high- and low-frequency image information as well as cloud regions in the training procedure. Our ablation and comparative experiments demonstrate that the DDPM-CR network outperforms other methods under various cloud conditions, achieving better visual effects and accuracy metrics (MAE = 0.0229, RMSE = 0.0268, PSNR = 31.7712, and SSIM = 0.9033). These results suggest that the DDPM-CR network is a promising solution for retrieving missing information in either thin or thick cloud-covered regions, especially when using auxiliary information such as SAR data. [ABSTRACT FROM AUTHOR]

NDVI is an important parameter for environmental assessment and precision agriculture that well-describes the status of vegetation. Nevertheless, the clouds in optical images often result in the absence of NDVI information at key growth stages. The integration of SAR and optical image features will likely address this issue. Although the mapping of different data sources is complex, the prosperity of deep learning technology provides an alternative approach. In this study, the double-attention RNN architecture based on the recurrent neural network (RNN) and attention mechanism is proposed to retrieve NDVI data of cloudy regions. Overall, the NDVI is retrieved by the proposed model from two aspects: the temporal domain and the pixel neighbor domain. The performance of the double-attention RNN is validated through different cloud coverage conditions, input ablation, and comparative experiments with various methods. The results conclude that a high retrieval accuracy is guaranteed by the proposed model, even under high cloud coverage conditions (R2 = 0.856, RMSE = 0.124). Using SAR images independently results in poor NDVI retrieval results (R2 = 0.728, RMSE = 0.141) with considerable artifacts, which need to be addressed with auxiliary data, such as IDM features. Temporal and pixel neighbor features play an important role in improving the accuracy of NDVI retrieval (R2 = 0.894, RMSE = 0.096). For the missing values of NDVI data caused by cloud coverage, the double-attention RNN proposed in this study provides a potential solution for information recovery. [ABSTRACT FROM AUTHOR]

With the gradual adjustment of urban expansion and discontinuation of non-essential functions in Beijing, many decades-old buildings have been demolished. Thus, construction and demolition waste (CDW) has become the focus of urban and dust pollution management. However, CDW piles are volatile and present irregular boundaries. Therefore, it is essential to map CDW regions in a timely and accurate manner to achieve urban development while protecting the environment. To address this issue, we proposed a method of CDW identification based on change detection and deep learning. First, ZY-3 multispectral images from 2016 and 2019 and their difference images were used for initial sample preparation. We expanded the samples using the post-classification comparison method of change detection, resulting in a 25.4% increase in the valid sample set. The expanded samples were then used as an input to the DeepLabV3+ for training. Thereafter, combined with the change information from the digital elevation model, specific forms, such as demolition remains, landfill CDW, and large-scale dump, were extracted using spatial analysis methods. The overall accuracy of CDW recognition was 91.67%, with a Kappa coefficient of 0.8642. In addition, we calculated the accuracy indices using only the initial samples, obtaining a mean Intersection-over-Union value that was 0.086 lower than that obtained using the expanded sample set. Similar results were obtained in PSPNet and UNet. This suggests that change detection is useful in improving the accuracy of the deep learning models. This study is the first to identify three existing forms of CDW and can effectively address the misclassification between CDW and bare land to identify CDW efficiently. [ABSTRACT FROM AUTHOR]

Hyperspectral remote sensing can get spatially and spectrally continuous data simultaneously. However, the imaging equipment is usually expensive and complex, along with the low spatial resolution. In recent years, reconstruction of hyperspectral image by deep learning from the widely used low-cost, high spatial resolution RGB camera, has attracted extensive attention in many fields. However, most research is limited to three bands in the range of 400–700 nm, which greatly restrains its application in remote sensing. In this study, a more suitable for remote sensing multispectral to hyperspectral network (M2H-Net) is proposed, which can take many bands as input and output hyperspectral images with any number of bands within a wider spectral range (380–2500 nm). Its characteristics include adding residual connection on U-Net to reduce vanishing gradients; adding convolution combinations with different kernel sizes (1 × 1 and 3 × 3) to balance the spectral and spatial relationships. It is applied on images from different platforms (UAVs and Satellites), different imaging modes (frame and pushbroom) and different spectral response functions (narrow and wide bandwidth), and the results show that: 1) it has a very high accuracy of hyperspectral image reconstruction. The mean relative absolute error (MRAE) and root mean squared error (RMSE) are between 0.039 and 0.074 and 0.010–0.016, respectively, which are 69.2% and 41.2% lower than those of U-Net; 2) it has high efficiency with fast convergence (about 40 epochs) and stable performance. Compared with many algorithms won in the new trends in image restoration and enhancement (NTIRE) competition, M2H-Net ranked 7th in accuracy, but took less time (0.44 s); 3) it has strong generalization ability. Using the pre-trained M2H-Nets to reconstruct Cubert S185 and GF-5 hyperspectral images in different locations, different times and complex scenes, high accuracy (MRAE = 0.072, RMSE = 0.011) can still be obtained. This method is more suitable for remote sensing to meet the needs of multiple bands, spectrum width and complex scenes, thus provides the possibility to generate the global coverage hyperspectral imagery by using the massive in-orbit or historical archived multispectral images, which will not only greatly save the R&D and investment on hyperspectral imaging equipment, but also conduct data collection with higher efficiency and lower complexity. Due to the ability to reconstruct hyperspectral images in specified bands on demand, M2H-Net is also of great value in hyperspectral image processing, such as data compression, storage and transmission, etc. [ABSTRACT FROM AUTHOR]