A5103081433- Medical Image Segmentation Techniques
- Image and Object Detection Techniques
- Image Retrieval and Classification Techniques
- Mathematical and Theoretical Epidemiology and Ecology Models
- Image and Signal Denoising Methods
- AI in cancer detection
- Advanced Differential Equations and Dynamical Systems
- Leaf Properties and Growth Measurement
- Video Surveillance and Tracking Methods
- Smart Agriculture and AI
- Evolution and Genetic Dynamics
- Image Processing Techniques and Applications
- Advanced Image Fusion Techniques
- Brain Tumor Detection and Classification
- Advanced Neural Network Applications
- Complex Network Analysis Techniques
- Human Pose and Action Recognition
- Video Analysis and Summarization
- Spectroscopy and Chemometric Analyses
- Biosimilars and Bioanalytical Methods
- Statistical Methods and Bayesian Inference
- Greenhouse Technology and Climate Control
- Refrigeration and Air Conditioning Technologies
- Statistical Methods and Inference
- Sparse and Compressive Sensing Techniques
Taizhou University
2022-2025
Wenzhou Medical University
2024
Zhejiang University
2024
Zhejiang Taizhou Hospital
2024
Qilu Hospital of Shandong University
2022
East China University of Technology
2013-2021
Central South University
2020-2021
Guangdong Ocean University
2007-2013
Changsha Normal University
2010-2011
Hunan Normal University
2010-2011
Active contour model (ACM) has been a successful method for image segmentation. The existing ACMs poorly segment the images with intensity inhomogeneity or non-homogeneity, and results highly depend on initial position of contour. To overcome these disadvantages, we proposed fuzzy region-based active driven by weighting global local fitting energy, wherein propose region energy spatial information, which proved convex ensures segmentation independent initialization, to motivate an evolving...
Active contour models (ACMs) have been widely applied in the field of image segmentation. However, it is still very challenging to construct an efficient ACM segment images with intensity inhomogeneity. In this paper, a novel guided by global and local signed energy-based pressure force (GLSEPF) proposed. First, computing energy difference between inner outer energies evolution curve, (GSEPF) designed, which can improve robustness initial curves. Second, (LSEPF) introduced pixel-by-pixel...
Precise delineation of hepatic lesions from multi-phase MRI scans is paramount for the prompt diagnosis and treatment liver conditions. Conventional approaches typically emphasize either global or local features, often disregarding spatial interdependencies across modalities difficulties arising ambiguous tumor margins. To address these deficiencies, we propose AMHMF-Net, an advanced adaptive modality-channel-search Hadamard multimodal fusion network. This architecture leverages a dual...
Precise delineation of hepatic lesions from multi-phase MRI scans is paramount for the prompt diagnosis and treatment liver conditions. Conventional approaches typically emphasize either global or local features, often disregarding spatial interdependencies across modalities difficulties arising ambiguous tumor margins. To address these deficiencies, we propose AMHMF-Net, an advanced adaptive modality-channel-search Hadamard multimodal fusion network. This architecture leverages a dual...
Introduction Pepper leaf segmentation plays a pivotal role in monitoring pepper diseases across diverse backgrounds and ensuring healthy growth. However, existing Transformer-based methods grapple with computational inefficiency, excessive parameterization, inadequate utilization of edge information. Methods To address these challenges, this study introduces an Adaptive Multi-Scale MLP (AMS-MLP) framework. This framework integrates the Multi-Path Aggregation Module (MPAM) Context Relation...
This study presents a novel active contour model (ACM) driven by weighted global and local region-based signed pressure force (SPF) to segment images in the presence of intensity inhomogeneity noise. First, an adaptive SPF (GRSPF) function as driving centers is designed based on image information, which normalized update weights inner outer regions curve during iterations. Second, introducing absolute differences weighs regions, (LRSPF) similarly defined. Third, instead setting fixed force,...
The segmentation of pepper leaves from images is great significance for the accurate control leaf diseases. To address issue, we propose a bidirectional attention fusion network combing convolution neural (CNN) and Swin Transformer, called BAF-Net, to segment image. Specially, BAF-Net first uses multi-scale feature (MSFF) branch extract long-range dependencies by constructing cascaded Transformer-based CNN-based block, which based on U-shape architecture. Then, it full-scale (FSFF) enhance...
Abstract Segmenting the region of interest (ROI) from medical images is a fundamental but challenging task due to illumination change and imaging devices. Although many models based on local region‐based active contour model (LR‐ACM) are proposed deal with intensity inhomogeneity, it still difficult for global energy‐based ACM image information accurately extract ROI images. To solve this problem, study proposes novel localised by constructing gradient probability scores fuzzy k‐ nearest...
Pepper leaf segmentation plays a crucial role in monitoring pepper diseases various backgrounds and ensuring the healthy growth of peppers. However, existing transformer-based methods suffer from computational inefficiency, excessive parameterization, limited utilization edge information. To tackle these challenges, we propose an adaptive multi-scale MLP framework, named AMS-MLP, which combines multi-path aggregation module (MPAM) context relation mask (MCRD) to refine object boundaries...
Abstract This paper presents an active contour model driven by adaptive‐scale local‐energy signed pressure force (ALSPF) function based on bias correction for segmenting medical images with intensity inhomogeneity. Firstly, a local energy‐driven (LESPF) is designed as the driving to extract image features, which can effectively deal Secondly, avid selecting neighbourhood size in LESPF function, multi‐scale adaptive selection schema put forward adaptively choose window according degree of...
This paper presents a novel fuzzy region-based active contour model for image segmentation. By incorporating local patch-energy functional along each pixel of the evolving curve into fuzziness energy, we construct patch-based energy function without regurgitation term. Its purpose is not only to make evolve very stably periodical initialization during evolution but also reduce effect noise. In particular, in order reject minimal functional, utilize direct method calculate alterations instead...
Representing an object with multiple image fragments or patches for target tracking in a video has proved to be able maintain the spatial information. The major challenges visual are effectiveness and robustness. We propose robust fragments-based algorithm adaptive feature selection. best discriminate is used tracking, which can improve effectiveness. A set of likelihood images corresponding most discriminative features fused divide into some fragments, By weighting fragment background...
Based on the predator-prey (natural enemy-pest) system with general functional response and seasonal effect, an impulsive differential to model processes of periodically releasing natural enemies spraying pesticides at different fixed times for pest control is considered. The dynamics are analyzed by using Floquet theory comparison techniques. Sufficient conditions local asymptotic stability global prey-eradication periodic solution established, respectively, latter corresponding success...
In this paper a delayed predator-prey model system with Holling-type IV functional response is studied. The bifurcation analysis of the shows that sequence Hopf bifurcations can occur at coexisting equilibrium as time delay crosses some critical values. An explicit algorithm for determining direction and stability bifurcating non-trivial periodic solutions derived by using normal form theory center manifold arguments due to Faria and. Finally, numerical simulations are carried out...
Abstract Background To evaluate the efficacy and safety of HLX01, a rituximab biosimilar, as combination therapy with methotrexate in Chinese patients active rheumatoid arthritis who had inadequate responses to methotrexate. Methods In this double-blind, placebo-controlled phase 3 trial, biologic-naïve moderate-to-severe were randomized 2:1 receive 1000 mg HLX01 or placebo intravenously on days 1 15. On first day weeks 24 26, both groups received via intravenous infusion. The primary...