A5003856741- Remote-Sensing Image Classification
- Infrared Target Detection Methodologies
- Remote Sensing and Land Use
- Optical Systems and Laser Technology
- Optical Polarization and Ellipsometry
- Spectroscopy and Chemometric Analyses
- Advanced Measurement and Detection Methods
- Advanced Image Fusion Techniques
- Image Processing Techniques and Applications
- Spectroscopy Techniques in Biomedical and Chemical Research
- Calibration and Measurement Techniques
- EEG and Brain-Computer Interfaces
- Image and Signal Denoising Methods
- Optical and Acousto-Optic Technologies
- Water Quality Monitoring and Analysis
- Advanced Image Processing Techniques
- Advanced Vision and Imaging
- Photoacoustic and Ultrasonic Imaging
- Advanced optical system design
- Gaze Tracking and Assistive Technology
- Functional Brain Connectivity Studies
- Plasmonic and Surface Plasmon Research
- Generative Adversarial Networks and Image Synthesis
- Image Enhancement Techniques
- Photonic and Optical Devices
Xi'an Institute of Optics and Precision Mechanics
2016-2025
Chinese Academy of Sciences
2016-2025
Spectral Imaging Laboratory (United States)
2020-2025
University of Chinese Academy of Sciences
2020-2024
Xi’an University
2020-2024
China XD Group (China)
2020
Shanghai Institute of Optics and Fine Mechanics
2015
Chinese Academy of Social Sciences
2013
Qingdao University
2009
In recent years, ship detection in satellite remote sensing images has become an important research topic. Most existing methods detect ships by using a rectangular bounding box but do not perform segmentation down to the pixel level. This paper proposes and method based on improved Mask R-CNN model. Our proposed can accurately segment at By adding bottom-up structure FPN of R-CNN, path between lower layers topmost layer is shortened, allowing features be more effectively utilized top layer....
Unmanned aerial vehicle (UAV) hyperspectral remote sensing technologies have unique advantages in high-precision quantitative analysis of non-contact water surface source concentration. Improving the accuracy non-point detection is a difficult engineering problem. To facilitate sensing, imaging, and spectral activities, UAV-based imaging system was designed. Its prototype built, laboratory calibration joint air–ground quality monitoring activity were performed. The UAV comprised light small...
A novel polarimetric dehazing method is proposed based on three linear polarization images (0°, 45°, and 90°). The orientation angle of the light scattered by haze particles introduced in algorithm. No additional image-processing algorithm needed postprocessing. It found that dehazed image suffers from little noise details objects close to observer can be preserved well. In addition, this also proved useful for preserving colors. Experimental results demonstrate such an has some universality...
In order to explore the application of hyperspectral technology in pathological diagnosis tumor tissue, we used microscopic imaging establish a database 30 patients with gastric cancer. Based on difference spectral-spatial features between cancer tissue and normal wavelength 410-910 nm, propose deep-learning model-based analysis method for tissue. The feature individual spatial-spectral joint medical contact are studied. experimental results show that classification accuracy proposed model...
Human pose estimation refers to accurately estimating the position of human body from a single RGB image and detecting location body. It serves as basis for several computer vision tasks, such tracking, 3D reconstruction, autonomous driving. Improving accuracy has significant implications advancement vision. This paper addresses limitations single-branch networks in estimation. presents top-down single-target approach based on multi-branch self-calibrating combined with graph convolutional...
With the development of industrialization and urbanization, consumption pollution water resources are becoming more serious. Water quality monitoring is an extremely important technical means to protect resources. However, current popular methods have their shortcomings, such as a low signal-to-noise ratio satellites, poor time continuity unmanned aerial vehicles, frequent maintenance in situ underwater probes. A non-contact near-surface system that can continuously monitor fluctuation...
Infectious diseases have always been a major threat to the survival of humanity. Additionally, they bring an enormous economic burden society. The conventional methods for bacteria identification are expensive, time-consuming and laborious. Therefore, it is great importance automatically rapidly identify pathogenic in short time. Here, we constructed AI-assisted system automating rapid genus identification, combining hyperspectral microscopic technology deep-learning-based algorithm Buffer...
Identifying infectious pathogens quickly and accurately is significant for patients doctors. single bacterial strains in eliminating culture speeding up diagnosis. We present an advanced optical method the rapid detection of (including common uncommon) by combining hyperspectral microscopic imaging deep learning. To acquire more information regarding pathogens, we developed a system with wide wavelength range fine spectral resolution. Furthermore, end-to-end learning network based on feature...
To achieve large-scale stitching of the hyperspectral remote sensing images obtained by unmanned aerial vehicles (UAVs) equipped with an acousto-optic tunable filter spectrometer, this study proposes a method based on feature fusion strategy and seam-finding using image classification. In extraction stage, SuperPoint deep features from in different spectral segments data cube were extracted fused. The depth matcher, LightGlue, was employed for matching. During unsupervised K-means...
Water, as a finite and vital resource, necessitates water quality monitoring to ensure its sustainable use. A key aspect of this process is the accurate measurement critical parameters such chemical oxygen demand (COD). However, current spectroscopic methods struggle with accurately consistently measuring COD in large-scale, complex environments due an insufficient understanding spectra limited generalizability. To address these limitations, we introduce physicochemical-informed spectral...
Multi-angle polarization characteristics of water-leaving radiation are crucial for the retrieval oceanic constituents and inherent optical properties (IOPs). However, feasibility using spectrum to retrieve IOPs in global oceans remains unclear. In this study, radiative transfer (RT) simulations Stokes vectors radiance (Lw) were performed, a polarization-based algorithm fully connected U-Net (FCUN) retrieving components was developed. The performance FCUN compared with situ measurements...
Compared with traditional single-view and radiometric-only observations, multiangle polarimetric observations of water-leaving radiation play a crucial role in enhancing the retrieval ocean constituents aerosol microphysical properties. In this study, impacts uncertainties degree polarization (DOP) radiance (Lw) on oceanic inherent optical properties (IOPs) were investigated via global radiative transfer (RT) simulations fully connected U-Net (FCUN) model. The IOPs further various sensor...
Abstract Objective. There is a notable need of quantifiable and objective methods for the classification schizophrenia. Patients with schizophrenia exhibit atypical eye movements compared healthy individuals. To address this need, we have developed model based on eye-tracking data to assist physicians in intelligent auxiliary diagnosis Approach. This study employed three experiments—Picture-Free Viewing, Smooth Pursuit Tracking, Fixation Stability—to collect from patients controls. The 292...