A5108856555- Medical Image Segmentation Techniques
- Gaze Tracking and Assistive Technology
- Power Systems Fault Detection
- Medical Imaging and Analysis
- Intermetallics and Advanced Alloy Properties
- Advanced Computing and Algorithms
- Advanced Image Fusion Techniques
- HVDC Systems and Fault Protection
- Power Systems and Technologies
- Advanced Computational Techniques and Applications
- Fuel Cells and Related Materials
- Smart Grid and Power Systems
- Aluminum Alloys Composites Properties
- Advanced Sensor and Energy Harvesting Materials
- High-Voltage Power Transmission Systems
- Electrocatalysts for Energy Conversion
- Optical Systems and Laser Technology
- Advancements in Solid Oxide Fuel Cells
- Advanced Neural Network Applications
- Advanced Decision-Making Techniques
- Robotics and Sensor-Based Localization
- Advanced Algorithms and Applications
- Infrared Target Detection Methodologies
- Advanced ceramic materials synthesis
- Tactile and Sensory Interactions
Chongqing University
2024
Shenyang Ligong University
2024
Wuhan Institute of Technology
2021-2023
California Institute for Biomedical Research
2023
Sichuan University of Science and Engineering
2012-2022
Capital Medical University
2012-2013
Capital University
2012
Hunan University
2002
Abstract Background Clinical diagnosis and therapy for the lumbar disc herniation requires accurate vertebra segmentation. The complex anatomical structure degenerative deformations of vertebrae makes its segmentation challenging. Methods An improved level set method, namely edge- region-based method (ERBLS), is proposed CT images By considering gradient information local region characteristics images, model can efficiently segment with intensity inhomogeneity blurry or discontinuous...
The fuel cell with a ten-channel serpentine flow field has low operating pressure drop, which is conducive to extended test operations and stable use. According numerical results of the cell, multi-channel usually poor mass transmission under ribs, lower drop not favorable for drainage from outlet. In this paper, an optimized developed address these two disadvantages cell. As per simulation, improves gas distribution in reaction area, increases between adjacent performance PEMFC, enhances...
Abstract In recent years, convolutional neural networks (CNNs) have been successfully applied to reconstruct images from speckle patterns generated as objects pass through scattering media. To achieve this objective, a large amount of data is collected for training the CNN. However, in certain cases, characteristics light passing medium may vary. such situations, it necessary collect substantial new re-train CNN and image reconstruction. address challenge, transfer learning techniques are...
Transmission lines are most prone to faults in the transmission system, so high-precision fault diagnosis is very important for quick troubleshooting. There some problems current intelligent research methods, such as difficulty extracting features accurately, low recognition accuracy and poor tolerance. In order solve these problems, this paper proposes an method high voltage direct (HVDC) based on Gramian angular difference field (GADF) domain improved convolutional neural network (VGG16)....
In brazes of pure nickel with Ni-Si-B filler metals, boride forms in the central region brazed gap by solidification, and base metal along bothjointfaying surfaces solid state diffusion precipitation. During subsequent heat treatment, gradually decomposes dissolves, while lining joint faying increases both size quantity. This phenomenon is result an uphill boron from gap, which remains very much a ternary system, into metal, largely binary Ni-B system owing to rapid ahead silicon metal. The...
The complex distribution, mutual occlusion, and scale difference greatly increase the difficulty of cotton detection in wild. To reduce omission ratio raise accuracy cotton, a dual-path feature extraction (DPFE) algorithm is proposed. It consists DPFE convolutional neural network, multi-path fusion module, multi-scale prediction module. First, uses Darknet network as main path for extraction. At same time, double downsampling map enhanced by proposed enhancement module—spatial pyramid...
Methods：In this paper, a new anatomic contour extraction algorithm based on the contourlet transform and pulse-coupled neural network ( PCNN ) was presented. The is applied vertebra CT image firstly, using its multiscale multidirectional characteristics to get coefficients which can effectively capture contours of an depict sparsely. Then utilized extract from approximation image. Finally, are reconstructed by inverse transform.Results：The proposed tested images, we some encouraging results,...
Fault phase selection is studied in this paper based on the wavelet power spectrum entropy and LIBSVM for transmission line. First, MATLAB used to simulate five fault types obtain a variety of voltage waveforms. Second, simulation waveform transformed extract energy which pretreatment greatly reduce number data dimensions effectively optimize input LIBSVM. Finally, select phase, experimental results show that method very effective.
In this paper, based on the studying of rank dioid matrix, some characteristics and laws matrix are discussed. Meanwhile, it is applied to equivalent system simplification model for DEDS.