A5024291763- Remote Sensing and Land Use
- Remote-Sensing Image Classification
- Cryospheric studies and observations
- Climate change and permafrost
- Remote Sensing in Agriculture
- Environmental Changes in China
- Arctic and Antarctic ice dynamics
- Land Use and Ecosystem Services
- Advanced Image and Video Retrieval Techniques
- Advanced Image Fusion Techniques
- Automated Road and Building Extraction
- Medical Image Segmentation Techniques
- Hydrology and Watershed Management Studies
- Infrared Target Detection Methodologies
- Advanced Computational Techniques and Applications
- Remote Sensing and LiDAR Applications
- Environmental and Agricultural Sciences
- Urban Heat Island Mitigation
- EEG and Brain-Computer Interfaces
- Image Retrieval and Classification Techniques
- Geographic Information Systems Studies
- Domain Adaptation and Few-Shot Learning
- Image and Signal Denoising Methods
- Meteorological Phenomena and Simulations
- Climate variability and models
Nanjing University
2016-2025
Ministry of Natural Resources
2020-2025
Southeast University
2023-2025
Shanghai Municipal Center For Disease Control Prevention
2024-2025
Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application
2016-2024
Shandong First Medical University
2024
Shandong Provincial Hospital
2024
State Key Laboratory of Digital Medical Engineering
2023
South China Institute of Collaborative Innovation
2016-2019
New York University Press
2018
Abstract Long-term datasets of number and size lakes over the Tibetan Plateau (TP) are among most critical components for better understanding interactions cryosphere, hydrosphere, atmosphere at regional global scales. Due to harsh environment scarcity data TP, accumulation sharing become more valuable scientists worldwide make new discoveries in this region. This paper, first time, presents a comprehensive freely available set lakes’ status (name, location, shape, area, perimeter, etc.) TP...
A critical obstacle to achieve semantic segmentation of remote sensing images by the deep convolutional neural network is requirement huge pixel-level labels. Taking building extraction as an example, this study focuses on how effectively apply weakly supervised (WSSS) high-resolution (HR) with image-level labels, which a prominent solution for labeling challenge. The widely used two-step WSSS framework adopted, in pseudo-masks are first produced from labels and followed trained...
Self-supervised learning (SSL) has gained widespread attention in the remote sensing (RS) and earth observation (EO) communities owing to its ability learn task-agnostic representations without human-annotated labels. Nevertheless, most existing RS SSL methods are limited either global semantic separable or local spatial perceptible representations. We argue that this strategy is suboptimal realm of RS, since required for different downstream tasks often varied complex. In study, we proposed...
Change detection is a critical task in earth observation applications. Recently, deep-learning-based methods have shown promising performance and are quickly adopted change detection. However, the widely used multiple encoders single decoder (MESD) as well dual-encoder–decoder (DED) architectures still struggle to effectively handle well. The former has problems of bitemporal feature interference feature-level fusion, while latter inapplicable intraclass (ICCD) multiview building (MVBCD). To...
China has experienced a rapid urban expansion over the past three decades because of its accelerated economic growth. In this study, we detected and analyzed during period using multi-temporal Defense Meteorological Satellite Program Operational Linescan System (DMSP-OLS) nighttime light data multi-source Normalized Difference Vegetation Index (NDVI) data. First, an intercalibration was performed to improve continuity comparability from 1992 2010. The NDVI were then subjected local support...
Learning effective visual representations without human supervision is a critical problem for the task of semantic segmentation remote sensing images (RSIs), where pixel-level annotations are difficult to obtain. Self-supervised learning (SSL), which learns useful by creating artificial supervised problems, has recently emerged as an method learn from unlabelled data. Current SSL methods generally trained on ImageNet through image-level prediction tasks. We argue that this suboptimal...
Semantic segmentation of remote sensing images is effective for large-scale land cover mapping, which heavily relies on a large amount training data with laborious pixel-level labeling. Weakly supervised semantic (WSSS) based image-level labels has attracted intensive attention due to its easy availability. However, existing WSSS methods mainly focus binary segmentation, are difficult apply multiclass scenarios. This study proposes comprehensive framework images, consisting appropriate label...
This paper presents a cosegmentation-based method for building change detection from multitemporal high-resolution (HR) remotely sensed images, providing new solution to object-based (OBCD). First, the magnitude of difference image is calculated represent feature. Next, cosegmentation performed via graph-based energy minimization by combining feature with features at each phase, directly resulting in foreground as changed objects and background unchanged area. Finally, spatial correspondence...
Snow cover in mountain areas is a key factor controlling regional energy balances, hydrological cycle, and water utilization. Optical remote sensing data offer an effective means of mapping snow cover, although their application limited by solar illumination conditions, conversely, synthetic aperture radar (SAR) offers the ability to measure wetness changes all weather. In this study, novel method, which can be approached two steps using SAR optical data, has been developed for dry wet...
China has undergone significant land cover changes since the 1980s. However, there are limited consistent and continuous dataset of national scale. Using advanced very high resolution radiometer moderate imaging spectrometer data from long-term record, we developed a large-scale classification approach to produce decadal 5-km for (ChinaLC) 1981 2010. A total 19 classes training validation samples were obtained visual interpretation high-resolution Google Earth images historical vegetation...
Abstract Extensive and complex changes in spring vegetation phenology have occurred the Pan‐Arctic over last several decades. However, role of snow cover at start growing season (SOS) under different climatic conditions remains unclear. Therefore, we compare effects four indicators on SOS from 1982 to 2015 based long‐term remote sensing data found that end date (SCED) is main indicator affecting SOS, with advancing 0.56 days for each 1‐day advance SCED, explaining 12%–90% variability 63%...