A5100448847- Advanced Image Processing Techniques
- Advanced Image Fusion Techniques
- Advanced Vision and Imaging
- Image and Signal Denoising Methods
- Remote-Sensing Image Classification
- Advanced Image and Video Retrieval Techniques
- Image Processing Techniques and Applications
- Infrared Target Detection Methodologies
- Sparse and Compressive Sensing Techniques
- Satellite Image Processing and Photogrammetry
- Face and Expression Recognition
- Medical Image Segmentation Techniques
- Image Enhancement Techniques
- Image Retrieval and Classification Techniques
- Remote Sensing in Agriculture
- Robotics and Sensor-Based Localization
- Digital Media Forensic Detection
- Remote Sensing and Land Use
- Spectroscopy and Chemometric Analyses
- Advanced Steganography and Watermarking Techniques
- Remote Sensing and LiDAR Applications
- Advanced optical system design
- Advanced Optical Imaging Technologies
- Microgrid Control and Optimization
- Radio Astronomy Observations and Technology
China Southern Power Grid (China)
2012-2025
Changsha University of Science and Technology
2010-2024
China Academy of Space Technology
2015-2024
The University of Texas at Tyler
2024
Michigan State University
2024
Universiti Teknologi MARA
2023
Meta (United States)
2023
Hebei Finance University
2023
Northeast Normal University
2023
BC Platforms (Finland)
2023
In this paper, we propose a new super resolution (SR) method called the maximum posteriori based on universal Hidden Markov Tree (HMT) model for remote sensing images. The HMT theory is first used to set up prior reconstructing resolved images from sequence of warped, blurred, subsampled, and noise-contaminated low-resolution (LR) Because wavelet coefficients can be well characterized as mixed Gaussian distribution, an better able capture dependences between multiscale coefficients. tested...
Visual tracking plays an important role in many computer vision tasks. A common assumption previous methods is that the video frames are blur free. In reality, motion blurs pervasive real videos. this paper we present a novel BLUr-driven Tracker (BLUT) framework for motion-blurred targets. BLUT actively uses information from without performing debluring. Specifically, integrate problem with motion-from-blur under unified sparse approximation framework. We further use inferred by to guide...
This paper investigates the role that nonlinear camera response functions (CRFs) have on image deblurring. We present a comprehensive study to analyze effects of CRFs motion In particular, we show how can cause spatially invariant blur behave as varying blur. prove such nonlinearity large errors around edges when directly applying deconvolution blurred without CRF correction. These are inevitable even with known point spread function (PSF) and state-of-the-art regularization-based...
In this paper, an unmanned aerial vehicle (UAV)-enabledspace-air-ground integrated relay system, in which UAV is equipped with phased-array antennas to receive satellite's signal while utilizing non-orthogonal multiple access (NOMA) schemes transmit data land-based users, proposed. For collection, a detection-vector optimization problem can be addressed by introducing serial of transformations into second-order cone programming problem. Furthermore, the UAV-ground NOMA communication...
Random microlens arrays (rMLAs) have been widely applied as a beam-shaping component within an optical system. Silica glass is undoubtedly the best choice for rMLAs because of its wide range spectra with high transmission and damage threshold. Yet, machining silica user-defined shapes still challenging. In this work, novel design fabrication methods random silica-glass (rSMLAs) are proposed detailed investigation technology presented. Based on molding technology, fabricated rSMLAs tunable...
We present a hybrid camera that combines the advantages of high resolution and speed camera. Our consists pair low-resolution high-speed (LRHS) cameras single high-resolution low-speed (HRLS) The LRHS are able to capture fast-motion with little motion blur. They also form stereo provide depth map. HRLS provides spatial but introduces severe blur when capturing fast moving objects. develop efficient algorithms simultaneously motion-deblur image reconstruct method estimates flow in then warps...
Filament eruptions are considered to be a common phenomenon on the Sun and other stars, yet they rarely directly imaged in meter decimeter wavebands. Using imaging data from DAocheng solar Radio Telescope (DART) 150--450 MHz frequency range, we present two eruptive filaments that manifest as radio dimmings (i.e., emission depressions). Simultaneously, portions of these discernible dark features chromospheric images. The sun-as-a-star flux curves brightness temperature, derived DART images,...
In this letter, the application of super-resolution (SR) techniques to GaoFen(GF)-4, which is most advanced geostationary-orbit earth observing satellite in China, remote sensing images investigated and tested. One shortcomings geostationary-orbit-based limitation spatial resolution. However, human beings never stop pursuing higher resolution images. This first experiment applying SR a sequence low-resolution (LR) captured by GF-4 within short time period. barriers for large gaps between...
This paper presents a novel coprime blurred pair (CBP) model to improve data security in camera surveillance. While most previous approaches have focused on completely encrypting the video stream, we introduce spatial encryption scheme by strategically blurring image/video contents. Specifically, form public stream and private original using two different kernels. Each will provide user who has lower clearance less access personally identifiable details while still allowing behavior be...
In this paper, a new framework of mixed sparse representations (MSRs) is proposed for solving ill-conditioned problems with remote sensing images. general, it very difficult to find common representation images because complicated ground features. Here we regard image as combination subimage smooth, edges, and point-like components, respectively. Since each domain transformation method capable representing only particular kind object or texture, group transformations are used sparsely...
Within the context of video frame interpolation, complex motion modeling is task capturing, in a sequence, where moving objects are located interpolated frame, and how to maintain temporal consistency motion. Existing interpolation methods typically assign either fixed size kernel or refined optical flow model motions. However, they have limitation data redundancy inaccuracy representation This paper introduces unified warping framework, named multi-scale expandable deformable convolution...
In this letter, the application of superresolution (SR) techniques to multispectral image clustering and classification is investigated tested using satellite data. A set images with better spatial resolution obtained after an SR technique applied several data sets recorded within a short period over study area. Improved performance demonstrated visually quantitatively by comparison original low-resolution or enlarged conventional interpolation method. This letter illustrates possibility...
The ability to quickly and intuitively edit digital content has become increasingly important in our everyday life. However, existing propagation methods for editing images are typically based on optimization with high computational cost large inputs. Moreover, generally inefficient highly time-consuming. Accordingly, improve efficiency, this paper proposes a novel method using bilateral grid, which can achieve instant of sparse image edits. Firstly, given an input user interactions, we...
Microscopic studies on thin film superconductors play an important role for probing non-equilibrium phase transitions and revealing dynamics at the nanoscale. However, magnetic sensors with nanometer scale spatial picosecond temporal resolution are essential exploring these. Here, we present all-optical, microwave-free method that utilizes negatively charged nitrogen-vacancy (NV) center in diamond as a non-invasive quantum sensor enables detection of Meissner state superconducting film. We...
We present a novel stereo imaging technique called dual-focus or DFSI. DFSI uses pair of images captured from different viewpoints and at foci, but with identical wide aperture size. Each image in an exhibits defocus blur, the two form defocused pair. To model we introduce kernel map (DKM) that computes size blur disk each pixel. derive disparity constraint for computing DKM DFSI, integrate estimation to simultaneously recover both maps. show recovered DKMs provide useful guidance segmenting...
The spatial resolution of an observation from a geostationary orbiting satellite is usually too coarse to track small scale macroalgae blooms. For mapping benefit orbit’s staring monitoring and frequent revisit intervals, we introduced super-resolution method that reconstructs high-resolution (HR) image region sequence raw low-resolution images the same region. We tested our with GF-4 at 50 m demonstrated increased 25 m. In addition, derived HR had better quality characterized by higher...