A5000376902- Remote-Sensing Image Classification
- Remote Sensing and Land Use
- Land Use and Ecosystem Services
- Data Management and Algorithms
- Geographic Information Systems Studies
- Constraint Satisfaction and Optimization
- Remote Sensing in Agriculture
- Automated Road and Building Extraction
- Impact of Light on Environment and Health
- Remote Sensing and LiDAR Applications
- Human Mobility and Location-Based Analysis
- Urban Transport and Accessibility
- Urban Heat Island Mitigation
- Advanced Image and Video Retrieval Techniques
- Advanced Computational Techniques and Applications
- Urban Green Space and Health
- Rough Sets and Fuzzy Logic
- Advanced Image Fusion Techniques
- Aeolian processes and effects
- Environmental Changes in China
- Image Retrieval and Classification Techniques
- Geochemistry and Geologic Mapping
- Species Distribution and Climate Change
- Urban Design and Spatial Analysis
- Soil erosion and sediment transport
Peking University
2016-2025
Hunan Normal University
2022-2023
State Key Laboratory of Remote Sensing Science
2007-2022
Hunan Provincial People's Hospital
2022
Ministry of Natural Resources
2021
Shanghai Jiao Tong University
2019
Center For Remote Sensing (United States)
2013-2014
Urumqi General Hospital of PLA
2013
Sun Yat-sen University
2008
South Central Minzu University
2007
In this paper, we propose a spectral–spatial feature based classification (SSFC) framework that jointly uses dimension reduction and deep learning techniques for spectral spatial extraction, respectively. framework, balanced local discriminant embedding algorithm is proposed extraction from high-dimensional hyperspectral data sets. the meantime, convolutional neural network utilized to automatically find spatial-related features at high levels. Then, fusion extracted by stacking together....
Timely and accurate classification interpretation of high-resolution images are very important for urban planning disaster rescue. However, as spatial resolution gets finer, it is increasingly difficult to recognize complex patterns in remote sensing images. Deep learning offers an efficient strategy fill the gap between image their semantic labels. due hierarchical abstract nature deep methods, capture precise outline different objects at pixel level. To further reduce this problem, we...
For hyperspectral remote sensing image (HSI) classification, the learning process of deep neural networks has been progressively advanced in depth, but fine features are often largely lost or even disappear depth transfer. With increase feature aggregation and connectivity, complexity network training parameters increases greatly, requiring more time. This paper proposed a multi-scale dense (MSDN) for HSI classification that made full use different scale information structure combined...
The sudden outbreak of the Coronavirus disease (COVID-19) swept across world in early 2020, triggering lockdowns several billion people many countries, including China, Spain, India, U.K., Italy, France, Germany, Brazil, Russia, and U.S. transmission virus accelerated rapidly with most confirmed cases U.S., Brazil. In response to this national global emergency, NSF Spatiotemporal Innovation Center brought together a taskforce international researchers assembled implementation strategies...
Semantic segmentation of remote sensing images is an important but unsolved problem in the society. Advanced image semantic models, such as DeepLabv3+, have achieved astonishing performance for semantically labeling very high resolution (VHR) images. However, it difficult these models to capture precise outlines ground objects and explore context information that revealing relationships among optimizing results. Consequently, this study proposes a method VHR by incorporating deep learning...
Urban functional zones (UFZs) are important for urban sustainability and planning management, but UFZ maps rarely available up-to-date in developing countries due to frequent economic human activities rapid changes UFZs. Current methods have focused on mapping UFZs a small area with either remote sensing images or open social data, large-scale integrating these two types of data is still not be applied. In this study, novel approach by (RSIs) proposed. First, context-enabled image...
Although semantic segmentation models based on deep neural networks (DNNs) have achieved excellent results, generalizing well from one remote sensing dataset (source domain) to another with different acquisition conditions (target remains a major challenge. Many domain adaptation (DA) approaches been proposed address this problem. DA aims help DNNs learn generalizable representation space in which source and target domains similar feature distributions, but most of the existing difficulty...
The start of the growing season (SOS) is essential to track responses vegetation climate change. However, recent findings on whether SOS in middle-high latitudes Northern Hemisphere (NH) continued advance or reversed during global warming hiatus were not consistent. It necessary investigate causes this controversy and examine relationship between preseason temperature trends. To end, we first applied four widely used phenology extraction methods derive from GIMMS NDVI3g dataset then ensemble...
Although much efforts have been made to develop automatic methods for building extraction from very high-resolution (VHR) imagery during the past 30 years; with high performance are still unavailable due three issues: uncertainty of segmentation scales, selection effective features, and sample selection. In this study, by introducing GIS data, a parameter mining approach is proposed (1) mine information extraction, (2) detect changes buildings between VHR data. For first target, learning...
The non-uniformity of the relationships between urban temperature and landscape has attracted board attention. in areas is reflected spatial landscape’s heterogeneity difference socio-economic functions. former shown as differentiation land-cover, land-use, composition, configuration, while latter leads to intensity human activities population density, which are closely related with anthropogenic heat emission. Therefore, this study introduces functional zones (UFZs) express heterogeneity....
Urban functional zones, such as commercial, residential, and industrial are basic units of urban planning, play an important role in monitoring urbanization. However, historical functional-zone maps rarely available for cities developing countries, traditional investigations focus on geographic objects rather than zones. Recent studies have sought to extract zones automatically from very-high-resolution (VHR) satellite images, they mainly concentrate classification techniques, but ignore...