Deep convolutional neural networks have become an indispensable method in remote sensing image scene classification because of their powerful feature extraction capabilities. However, the ability of the models to extract multi-scale features and global features on surface objects of complex scenes is currently insufficient. We propose a framework based on global context spatial attention (GCSA) and densely connected convolutional networks to extract multi-scale global scene features, called GCSANet. The mixup operation is used to enhance the spatial mixed data of remote sensing images, and the discrete sample space is rendered continuous to improve the smoothness in the neighborhood of the data space. The characteristics of multi-scale surface objects are extracted, and their internal dense connection is strengthened by the densely connected backbone network. GCSA is introduced into the densely connected backbone network to encode the context information of the remote sensing scene image into the local features. Experiments were performed on four remote sensing scene datasets to evaluate the performance of GCSANet. The GCSANet achieved the highest classification precision on AID and NWPU datasets and the second best performance on the UCM dataset, indicating the GCSANet can effectively extract the global features of remote sensing images. In addition, the GCSANet presents the highest classification accuracy on the constructed mountain image scene dataset. These results reveal that the GCSANet can effectively extract multi-scale global scene features on complex remote sensing scenes. The source codes of this method can be found in https://github.com/ShubingOuyangcug/GCSANet.
Owing to their powerful feature extraction capabilities, deep learning-based methods have achieved significant progress in hyperspectral remote sensing classification. However, several issues still exist in these methods, including a lack of hyperspectral datasets for specific complicated scenarios and the need to improve the classification accuracy of land cover with limited samples. Thus, to highlight and distinguish effective features, we propose a hyperspectral classification framework based on a joint channel-space attention mechanism and generative adversarial network (JAGAN). To relearn featurebased weights, a higher priority was assigned to important features, which was developed by integrating a two-joint channelspace attention model to obtain the most valuable feature via the attention weight map. Additionally, two classifiers were designed in JAGAN: sigmoid was used to determine whether the input data were real or fake samples produced by the generator, while Softmax was adopted as a land cover classifier to yield the prediction type labels of the input samples. To test the classification performance of the JAGAN model, we used a selfconstructed complex land cover dataset based on GaoFen-5 AHSI images, which consists of mixed landscapes of mining and agricultural areas from the urban-rural fringe. Compared with other methods, the proposed model achieved the highest overall classification accuracy of 86.09%, the highest kappa amount of 79.41%, the highest F1 score of 85.86%, and the highest average accuracy of 82.30%, indicating the JAGAN can effectively improve the classification accuracy for limited samples in complex regional environments using GF-5 AHSI images.
<div><span>With the rapid development of urbanization in China, urban circles and urban agglomerations are gradually formed among different cities, which in turn has brought large pressure to the ecological environment. As an important monitoring index for evaluating the sustainable development of cities, quantified evaluation on the eosystem health is lacked for urban agglomerations. In this study, ecosystem health was assessed based on the framework of ecosystem vigor, organization, resilience, and services (VORS) in the Middle Reaches of the Yangtze River Urban Agglomerations (MRYRUA) in 2000, 2005, 2010, and 2015 with county as research units. Using GeoDetector to quantitatively analyze the impact of seven factors (including the proportion of construction land, forest land, and water, land use degree, population, average annual precipitation, and digital elevation model (DEM)) on ecosystem health in different periods. The results showed that: (1) There were significant differences in the spatial distribution of ecosystem health. The ecosystem health in the central area of Wuhan Metropolis, Changsha-Zhuzhou-Xiangtan City Group, and Poyang Lake City Group were significantly lower than the surrounding areas; (2) From the time scale, the research units of ordinary well level gradually develop to relatively well and well levels. The research units of relatively weak and weak level remain relatively stable. (3) Land use degree was the main factor affecting on ecosystem health. Moreover, there were interactions between different factors affecting. The impact of factors on ecosystem health were bi-enhanced or nonlinear enhanced. (4) The impacts of the proportion of construction land on ecosystem health had become greater over the time, and risen from fourth in 2000 to second in 2015. Therefore, a reasonable layout of urban land use planning has an important impact on the ecosystem health.</span></div>
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.