A5100408266- Remote-Sensing Image Classification
- Advanced Image Fusion Techniques
- Remote Sensing and Land Use
- Engineering Education and Curriculum Development
- 3D Surveying and Cultural Heritage
- Chemistry and Chemical Engineering
- Remote Sensing and LiDAR Applications
- Remote Sensing in Agriculture
- Image Processing Techniques and Applications
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Aquatic Ecosystems and Phytoplankton Dynamics
- Satellite Image Processing and Photogrammetry
- ICT Impact and Policies
- Advanced Vision and Imaging
- Land Use and Ecosystem Services
- Automated Road and Building Extraction
- Marine and coastal ecosystems
- Image and Signal Denoising Methods
- Advanced Optical Imaging Technologies
- Medical Image Segmentation Techniques
- Engineering Education and Pedagogy
- Sparse and Compressive Sensing Techniques
- Optical Systems and Laser Technology
- Environmental Changes in China
- Aquatic Ecosystems and Biodiversity
Jiangxi Normal University
2010-2025
Tsinghua University
2006-2024
Hainan Maternal and Child Health Hospital
2024
Jiaozuo University
2024
Henan Polytechnic University
2024
Wuhan University
2015-2023
International Centre on Space Technologies for Natural and Cultural Heritage
2023
California University of Pennsylvania
2023
Xi'an High Tech University
2023
State Key Laboratory of Information Engineering in Surveying Mapping and Remote Sensing
2016-2022
Deep convolution neural network (DCNN) is among the most effective ways of performing land use classification high-resolution remote sensing images. Land by fusing optical and synthetic aperture radar (SAR) images has broad application prospects, but related research studies are few. In this study, we developed first largest joint SAR dataset, WHU-OPT-SAR, covering an area approximately 50,000 km2, designed a multimodal-cross attention (MCANet). MCANet comprises three core modules:...
Abstract In the valley of Bamiyan, Afghanistan, approximately 1700 years ago, two large standing Buddha statues were carved out sedimentary rock region. They 53 and 38 m high larger one figured as tallest representation a in world. March 2001 Taleban government militia demolished colossal statues. After destruction group from ETH Zürich completed computer reconstruction Great Buddha, which can serve basis for physical reconstruction. This paper reports results image‐based 3D statue,...
An effective similarity measure for multi-modal images is crucial medical image fusion in many clinical applications. The underlining correlation across modalities usually too complex to be modelled by intensity-based statistical metrics. Therefore, approaches of learning a metric are proposed recent years. In this work, we propose novel deep method that trains binary classifier learn the correspondence two patches. classification output transformed continuous probability value, then used as...
Traditional villages face many difficulties in the era of globalization, especially light fast industrialization and urbanization. The breakdown settlement patterns erosion local characteristics cultural identities pose critical issues for sustainable development these communities. While research on traditional dwellings core areas is relatively advanced, there remains a significant gap studies focusing multicultural intermingling regions. By clarifying landscapes under influence multiple...
The second batch of Zhuhai-1 microsatellites was successfully launched on 26 April 2018. included four Orbita hyperspectral satellites (referred to as OHS-A, OHS-B, OHS-C, and OHS-D) one video satellite (OVS-2A), which have excellent data acquisition abilities. For the first time in China, a number networks been realized. To ensure application remote sensing data, series on-orbit geometry processing accuracy verification studies has carried out “Zhuhai-1” camera since launched. This paper...
(2020). Evaluating generative adversarial networks based image-level domain transfer for multi-source remote sensing image segmentation and object detection. International Journal of Remote Sensing: Vol. 41, No. 19, pp. 7343-7367.
Abstract. Purpose: Different types of stereoscopic video displays have recently been introduced. We measured and compared visual fatigue discomfort induced by viewing two different that either used the pattern retarder‐spatial domain technology with linearly polarized three‐dimensional or circularly using shutter glasses. Methods: During this observational cross‐over study performed at subsequent days, a was watched 30 subjects (age: 20–30 years). Half participants screen retard display...
China’s largest freshwater lake, Poyang Lake, is characterized by rapid changes in its inundation area and hydrodynamics, so this study, a hydrodynamic model of Lake was established to simulate these long-term changes. Inundation information extracted from Moderate Resolution Imaging Spectroradiometer (MODIS) remote sensing data used calibrate the wetting drying parameter assessing accuracy simulated boundary. The bottom friction calibrated using current velocity measurements Acoustic...
The deterioration of water quality has become a primary environmental concern worldwide. Understanding the status and identifying influencing factors are important for resources management. However, reported analyses have mostly been conducted in small focused areas. It is still unclear if driving spatial variation would be different extended scales. In this paper, we analyzed pattern inland surface China using dataset with four parameters (i.e., pH, DO, NH4+-N CODMn) level. We tested...
China’s aging population has deeply affected the sustainable development of Chinese economy. Based on provincial panel data and economic indicators from 2000 to 2020, this paper develops a vector autoregressive model analyze effect growth under paths household consumption national savings, respectively. The results show that an inhibits promotes saving, which both direct indirect effects growth. In particular, is not conducive in context slow over past three years contraction global economy...
With the continuous improvement of China’s economic strength, employment problem college graduates is gradually highlighted. At present, adjustment industrial structure and change in demand for talent market cause to face severe pressure seriously impact their future career development. In this regard, paper first briefly analyzes from perspective high-quality development then puts forward effective countermeasures related problems promote smooth lay a solid foundation [1].
Although changes in water temperature influence the rates of many ecosystem processes lakes, knowledge regime for large floodplain lake systems subjected to multiple stressors has received little attention. The coupled models can serve derive more on impact ecosystems. For this purpose, we used a physically-based hydrodynamic model with transport examine spatial and temporal behavior primary causal factors within Poyang Lake that is representative shallow lakes China. Model performance...
Although mixing in lakes has significant environmental and ecological implications, knowledge of dynamics for shallow floodplain received little attention. In this study, hydrological, thermal isotopic investigations were undertaken to provide evidence the large, Poyang Lake (China). Depth profiles water velocity, temperature stable hydrogen oxygen isotope compositions measured throughout lake, with results showing that velocity differences depth are generally less than ∼0.2 m/s, indicating...
Hyperspectral image (HSI) classification has attracted widespread concern in recent years. However, due to the complexity of HSI gathering environment, it is difficult obtain a great number labeled samples. Therefore, how effectively extract spatial–spectral feature with small-scale training samples crucial point classification. In this paper, novel fusion framework for small-sample proposed fully combine advantages multidimensional CNN and handcrafted features. Firstly, 3D fuzzy histogram...
The automatic 3D reconstruction of indoor scenes is great significance in the application 3D-scene understanding. existing methods have poor resilience to incomplete and noisy point cloud, which leads low-quality results tedious post-processing. Therefore, objective this work automatically reconstruct from an point-cloud base on semantics primitives. In paper, we propose a semantics-and-primitives-guided method. Firstly, local, fully connected graph neural network designed for semantic...
Extracting buildings from remote sensing images plays an important role in urban development planning, disaster assessment and mapping. Convolutional neural network (CNN) has been widely applied to building extraction because of its powerful deep semantic feature ability. However, existing CNN-based methods are difficult accurately extract multiscale with accurate edges the limitation receptive fields loss spatial detail information. For above problems, this paper proposes a field (MSRF-Net)...
Most active scene recovery techniques assume that a point is illuminated only directly by the illumination source. Consequently, global effects due to inter-reflections, sub-surface scattering and volumetric introduce strong biases in recovered shape. Our goal recover properties presence of illumination. To this end, we study interplay between depth cue defocus. By expressing both these as low pass filters, derive an approximate invariant can be used separate them without explicitly modeling...