A5028584961- Remote-Sensing Image Classification
- Remote Sensing and Land Use
- Advanced Image Fusion Techniques
- Physics of Superconductivity and Magnetism
- Advanced Image Processing Techniques
- Image and Signal Denoising Methods
- Theoretical and Computational Physics
- Semiconductor materials and devices
- Semiconductor Quantum Structures and Devices
- Advanced Semiconductor Detectors and Materials
- GaN-based semiconductor devices and materials
- Advanced Vision and Imaging
- Land Use and Ecosystem Services
- Spectroscopy and Chemometric Analyses
- Electrocatalysts for Energy Conversion
- Remote Sensing in Agriculture
- Ultrasound Imaging and Elastography
- ZnO doping and properties
- Advanced Chemical Sensor Technologies
- Advanced Sensor and Energy Harvesting Materials
- Catalytic Processes in Materials Science
- Identification and Quantification in Food
- Advanced Image and Video Retrieval Techniques
- Sparse and Compressive Sensing Techniques
- Magnetic properties of thin films
Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou)
2020-2025
Sun Yat-sen University
2016-2025
Czech Academy of Sciences, Institute of Physics
2024
Huazhong Agricultural University
2024
Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai)
2020-2024
Hubei University of Automotive Technology
2024
Shandong First Medical University
2024
Academy of Medical Sciences
2024
Zhejiang Provincial Public Security Department
2024
Chinese Academy of Sciences
2004-2022
Mangroves are one of the most important coastal wetland ecosystems, and compositions distributions mangrove species essential for conservation restoration efforts. Many studies have explored this topic using remote sensing images that were obtained by satellite-borne airborne sensors, which known to be efficient monitoring ecosystem. With improvements in carrier platforms sensor technology, unmanned aerial vehicles (UAVs) with high-resolution hyperspectral both spectral spatial domains been...
Classification of hyperspectral image (HSI) is an important research topic in the remote sensing community. Significant efforts (e.g., deep learning) have been concentrated on this task. However, it still open issue to classify high-dimensional HSI with a limited number training samples. In paper, we propose semi-supervised classification method inspired by generative adversarial networks (GANs). Unlike supervised methods, proposed semi-supervised, which can make full use labeled samples as...
To accurately estimate leaf area index (LAI) in mangrove areas, the selection of appropriate models and predictor variables is critical. However, there a major challenge quantifying mapping LAI using multi-spectral sensors due to saturation effects traditional vegetation indices (VIs) for forests. WorldView-2 (WV2) imagery has proven be effective grasslands forests, but sensitivity its been uncertain Furthermore, single model may exhibit certain randomness instability calibration estimation...
Hyperspectral videos (HSVs) play an important role in the monitoring domain, as they can provide more information than RGB about movement of interesting objects from perspective material interpretation. However, acquisition HSV data is expensive and time-consuming, whereas are readily available. In order to obtain its corresponding data, this paper proposes a lightweight frequency-spectrum unfolding network (FSUF-Net) for spectral super-resolution (SSR) video data. Specifically, proposed...
Wetlands play a crucial role in achieving carbon peak and neutrality goals. Exploring spatiotemporal distribution is one of the fundamental task wetland research. However, existing large-scale long-term series mapping methods have challenges related to classification accuracy obtaining inter-annual samples. Therefore, rapid sample collection precise method are needed support resource assessment, conservation, ecological restoration. In this paper, we propose novel deep learning suitable for...
Remote sensing is a powerful technology for Earth observation (EO), and it plays an essential role in many applications, including environmental monitoring, precision agriculture, resource managing, urban characterization, disaster emergency response, etc. However, due to limitations the spectral, spatial, temporal resolution of EO sensors, there are situations which remote data cannot be fully exploited, particularly context response (i.e., applications real/near-real-time needed)....
Investigating mangrove species composition is a basic and important topic in wetland management conservation. This study aims to explore the potential of close-range hyperspectral imaging with snapshot sensor for identifying under field conditions. Specifically, we assessed data pre-processing transformation, waveband selection machine-learning techniques develop an optimal classification scheme eight Qi’ao Island Zhuhai, Guangdong, China. After five spectral datasets, which included...
With the development of social media, it has become increasingly important to quickly and accurately identify media texts related disasters (e.g. typhoon) aid in rescue recovery efforts. Currently, multi-class classification pre-trained language model Bidirectional Encoder Representations from Transformers (BERT) are widely used for text classification. However, most studies on typhoon damage single-label, which contradicts reality that a may correspond multiple types damage. Moreover,...
Recently, many researchers have attempted to exploit spectral–spatial features and sparsity-based hyperspectral image classifiers for higher classification accuracy. However, challenges remain efficient feature generation combination in the classifiers. This paper utilizes empirical mode decomposition (EMD) morphological wavelet transform (MWT) gain features, which can be significantly integrated by sparse multitask learning (MTL). In extraction step, sum of intrinsic functions extracted an...
Due to the capacity of full-time and full-weather working, synthetic aperture radar (SAR) images have been frequently applied ship detection. However, interference speckle noise shores poses enormous challenges accuracy Extracting multi-scale features is regarded as a good way detect ships different sizes, but at scales are not strictly aligned, which may further affect detection results. Therefore, this letter proposes an anchor-free method, namely power transformations feature alignment...
Osteoarthritis (OA) is a chronic joint disease highly associated with an imbalance in the network of inflammatory factors and typically characterized by oxidative stress cartilage damage. Moreover, specificity structure makes it difficult for drugs to achieve good penetration effective enrichment cavity. Therefore, therapeutic strategies that increase specific targeting incorporate antioxidative effects are important improve efficacy OA. Here, we developed folic acid-modified liposomal...
Multitask learning (MTL) has recently yielded impressive results for classification of remotely sensed data due to its ability incorporate shared information across multiple tasks. However, it remains a challenging issue achieve robust in the case that are from nonlinear subspaces. In this paper, we propose kernel low-rank MTL (KL-MTL) method handle features 2-D variational mode decomposition (2-D-VMD) domain multi-/hyperspectral classification. On one hand, nonrecursive 2-D-VMD is applied...
Existing hyperspectral sensors usually produce high-spectral-resolution but low-spatial-resolution images, and super-resolution has yielded impressive results in improving the resolution of images (HSIs). However, most methods require multiple observations same scene improve spatial without fully considering spectral information. In this paper, we propose an HSI method inspired by deep Laplacian pyramid network (LPN). First, is enhanced LPN, which can exploit knowledge from natural using any...
The ability to engineer microscale and nanoscale morphology upon metal nanowires (NWs) has been essential achieve new electronic photonic functions. Here, this study reports an optically programmable Plateau-Rayleigh instability (PRI) demonstrate a facile, scalable, high-resolution engineering of silver NWs (AgNWs) at temperatures <150 °C within 10 min. This accomplished by conjugating photosensitive diphenyliodonium nitrate with AgNWs modulate surface-atom diffusion. conjugation is...
Hyperspectral image processing is faced with difficulties considering its redundant features and complex information. Studies on hyperspectral feature extraction in the deep learning domain have become increasingly popular. The mainstream techniques fully consider spatial information local neighborhoods when extracting spectral by constructing neural networks. Deep generative models simulate intrinsic structure of samples adequately training, showing their potential values for signal...