A5066140811- Image and Signal Denoising Methods
- Advanced Image Processing Techniques
- Sparse and Compressive Sensing Techniques
- Advanced Image Fusion Techniques
- Remote-Sensing Image Classification
- Advanced Vision and Imaging
- Advanced Image and Video Retrieval Techniques
- Image Retrieval and Classification Techniques
- Pulmonary Hypertension Research and Treatments
- Plasma Diagnostics and Applications
- Occupational Health and Safety Research
- Remote Sensing and Land Use
- Protein Tyrosine Phosphatases
- Risk and Safety Analysis
- Human-Automation Interaction and Safety
- ATP Synthase and ATPases Research
- Robotics and Sensor-Based Localization
- Plasma Applications and Diagnostics
- Image and Object Detection Techniques
- Diamond and Carbon-based Materials Research
Nanchang Institute of Technology
2019-2024
Nanchang Institute of Science & Technology
2021-2023
Air Force Engineering University
2022
Abstract This study aims to improve impurity analysis by plasma electron spectroscopy for organic molecules. Various impurities can be registered simultaneously in one measurement, because the appearance energies of characteristic Penning electrons vary different chemical compounds. Herein, experimental studies were conducted on helium with alcohol vapor a nonlocal negative glow short micro-discharge an increase pressure from 15 Torr 150 Torr. As result, enables detection gas high-pressure...
In this paper, a new model for image restoration under Poisson noise based on high‐order total bounded variation is proposed. Existence and uniqueness of its solution are proved. To find the global optimal our strongly convex model, split Bregman algorithm introduced. Furthermore, rigorous convergence theory proposed established. Experimental results provided to demonstrate effectiveness efficiency method over classic variation‐based model.
Blind image deconvolution has attracted growing attention in processing and computer vision. The total variation (TV) regularization can effectively preserve edges. However, due to lack of self-adaptability, it does not perform very well on restoring images with complex structures. In this paper, we propose a new blind model using an adaptive weighted TV regularization. This better handle local features image. Numerically, design effective alternating direction method multipliers (ADMM)...
Many recent studies have shown that Euler’s elastica regularization performs better than the famous total variation (TV) on keeping image features in smooth regions during process of denoising. In addition, an adaptive weighted matrix combined with is a key technique for well restoring local image. Considering these two factors, this paper, we propose model Poisson restoration so as to preserve both and To solve non-smooth non-convex efficiently, design alternating direction method...
In this paper, a new relaxation model based on mean curvature for adaptive image restoration is proposed. To solve the problem efficiently, an alternating direction method of multipliers (ADMMs) Furthermore, rigorous convergence theory proposed algorithm established. We also give complexity analysis our method. Experimental results are provided to demonstrate effectiveness and efficiency over state-of-the-art synthetic natural images.
In the process of image stitching, many problems will inevitably arise, such as misalignment, artifacts and local structure distortion in overlapping regions. A parallax stitching algorithm combining improved feature optimization with an innovative iterative seam estimation is proposed. First, point features line input images are detected. To optimize features, histogram statistical approach proposed to remove false matching points combined RANSAC algorithm. Second, mesh warp optimized by...
The point features of low-texture images are insufficient and unreliable, so it is difficult to achieve good alignment easy damage the image structure. To solve these problems, in this paper, we propose a new stitching method by using sigmoid function create perception mask. Firstly, line used improve accuracy registration naturalness distortion. Secondly, an energy optimize model. Finally, use mask reduce artifacts retain gradient domain fusion algorithm combined fusion. Experimental...
The fusion of a low spatial resolution hyperspectral image (LR-HSI) and high multispectral (HR-MSI) in the same scene is common method to get (HR-HSI). For drawback that standard nuclear norm regularization treats each singular value equally, this paper proposes weighted model based on sparse matrix factorization (called WNNS) for fusion. Specifically, we promote sparsity fused images by adding ℓ<sub>1 </sub>norm coefficients. Furthermore, preserve important data components, combine with...
Hyperspectral images usually have higher spectral resolution but lower spatial resolution, compared with the multispectral images. Low brings difficulties to practical applications of hyperspectral Therefore, get high image, it is very important fuse low image in same scene. In this paper, we propose a hybrid regularization model by integrating sparse prior, local low-rank and total variation based on l <inf xmlns:mml="http://www.w3.org/1998/Math/MathML"...