A5029571917- Neuroscience and Neural Engineering
- Advanced Algorithms and Applications
- Advanced Measurement and Detection Methods
- Advanced Steganography and Watermarking Techniques
- Advanced Image and Video Retrieval Techniques
- Advanced Vision and Imaging
- Non-Destructive Testing Techniques
- Conducting polymers and applications
- Chaos-based Image/Signal Encryption
- Video Coding and Compression Technologies
- Advanced Memory and Neural Computing
- Advanced Sensor and Control Systems
- Infrared Target Detection Methodologies
- High voltage insulation and dielectric phenomena
- EEG and Brain-Computer Interfaces
- Power Transformer Diagnostics and Insulation
- Medical Image Segmentation Techniques
- Image Enhancement Techniques
- Advanced Data and IoT Technologies
- Electrical Contact Performance and Analysis
- Advanced Fiber Optic Sensors
- Photovoltaic System Optimization Techniques
- Digital Imaging for Blood Diseases
- Robotics and Sensor-Based Localization
- Power Line Inspection Robots
Shanghai University of Electric Power
2017-2025
PowerChina (China)
2025
Beijing Jiaotong University
2012-2013
Institute of Information Science
2013
Up to now, a watermarking scheme that is robust against desynchronization attacks (DAs) still grand challenge. Most image resynchronization schemes in literature can survive individual global DAs (e.g., rotation, scaling, translation, and other affine transforms), but few are resilient challenging cropping local DAs. The main reason features for watermark synchronization only globally invariable rather than locally invariable. In this paper, we present blind transform attacks. First, propose...
This paper investigates the disturbance observer based robust cruise control problem for heavy haul train with input saturation and disturbances. Both uncertain parameters actuator are taken into account in dynamic model of train. To reject influence disturbances from channel, a linear is proposed to approximate unknown disturbance, an augmented system constructed by combining state estimation error. According Lyapunov stability analysis method guaranteed cost theory, sufficient condition...
Snakes, or active contours, have been widely used in image processing applications. An external force for snakes called gradient vector flow (GVF) attempts to address traditional snake problems of initialization sensitivity and poor convergence concavities, while generalized GVF (GGVF) aims improve long thin indentations (LTIs). In this paper, we find show that both GGVF essentially yield the same performance capturing LTIs odd widths, generally neither can converge even-width LTIs. Based on...
Abstract In implantable electrophysiological recording systems, the headstage typically comprises neural probes that interface with brain tissue and integrated circuit chips for signal processing. While advancements in MEMS CMOS technology have significantly improved these components, their interconnection still relies on conventional printed boards sophisticated adapters. This approach adds considerable weight volume to package, especially high channel count systems. To address this issue,...
Wind speed fluctuation generally shortens the lifetime of insulated gate bipolar transistor (IGBT) module in doubly fed induction generator (DFIG), which leads to frequent component replacement and cost increase wind power generation. Conventional control strategies consider solely turbine output maximization, existing junction temperature thermal management focus on IGBT extension, both are not effective under turbulent conditions ignore trade-off between device reliability for overall DFIG...
In recent years, large-scale renewable energy access to substations has brought overload, harmonic, short circuit and other problems, which led an increase in the failure rate shortening service life of important power equipment such as transformers. Transformer is one key system, its operation status impact on safe stable grid. order realize real-time state evaluation transformer, a vibration signal detection method based video proposed this paper. Firstly, YOLOv4 used detect transformer...
Abstract In implantable electrophysiological recording systems, the headstage typically consists of neural probes interfacing with brain tissue and integrated circuit chips for signal processing. Although advancements in MEMS CMOS technology have significantly improved these components, their connection still relies heavily on conventional printed boards sophisticated adapters. This traditional approach adds considerable weight volume, especially as channel counts increase. To address this,...
To address the problems of low foreign object detection accuracy and high model complexity due to many small targets complex surrounding environment in transmission line scenarios, a network CFG-YOLOv8n is proposed based on YOLOv8n. Firstly, CA mechanism introduced increase attention YOLOv8n occluded targets, so as accuracy. Secondly, ghost convolution employed replace conventional backbone neck networks reduce parameter amount. Finally, Focal-EIoU loss alleviate sample imbalance problem...
Partial discharge (PD) is the main cause of insulation breakdown in high voltage electrical equipment, which poses a potential threat to safety and reliability equipment. Traditional PD detection methods have limitations liquid systems. In this paper, an interferometric method for detecting proposed. Through experimental platform based on Michelson interference method, water sample excited by applied lightning surge generator, fringes distorted are collected CCD camera. The results show that...
Wireless neural signal transmission is essential for both neuroscience research and disorder therapies. However, conventional wireless systems are often constrained by low sampling rates, limited channel counts, their support of only a single mode. Here, we developed bi-directional brain-computer interface system featuring dual modes. This supports low-power Bluetooth high-sampling-rate Wi-Fi transmission, providing flexibility various application scenarios. The mode, with maximum rate 14.4...
<title>Abstract</title> Local feature extraction and matching has lately attracted increasing attention due to its wide application, especially in real-time automated systems. However, existing image methods struggle balance the global receptive field efficient computation, which limits practical applications. Recently, State Space Model (SSM) shown great potential linear complexity long-range dependency modeling. Therefore, this paper, a local method using SSM is proposed, aims achieve...
The region of interest (ROI) segmentation infrared image is one the key steps in intelligent fault diagnosis power equipment.Subsequently, a two-dimensional entropy multilevel threshold based on bat algorithm proposed.By analyzing principle with entropy, used to search optimal threshold, and these thresholds are experiment equipment.The results show that this method improves time consuming precision compared multi-threshold particle swarm optimization algorithm, which effectively solves...
Infrared image segmentation of power equipment is the basis for intelligent diagnosis faults. In order to reduce influence non-uniform background on infrared and improve accuracy efficiency segmentation, an improved Niblack method based bat algorithm proposed. This uses variance between classes as fitness function automatically search optimal threshold non-overlapping rectangular neighborhood in method, it binarization current extract target area from image. Experimental results show that...
With the rapid development of multi-view video coding (MVC) and light field rendering (LFR), Free-View Television (FTV) has emerged as new entrainment equipment, which can bring more immersive realistic feelings for TV viewers. In FTV broadcasting system, TV-viewer freely watch a arbitrary view scene generated from number original views. such scenario, ownership should be verified not only on views, but also any virtual view. However, capacities existing watermarking schemes copyright...
<title>Abstract</title> Local feature detection and description play a crucial role in various computer vision tasks, including image matching. Variations illumination conditions significantly affect the accuracy of these applications. However, existing methods inadequately address this issue. In paper, novel algorithm based on Illumination Auxiliary Learning Module (IALM) is introduced. Firstly, new local extractor named SuperPoint (IA-SuperPoint) established, integration IALM SuperPoint....
A novel multi-motor drives, based on series-connected PMSM, has been proposed recently for the requirement of superpower and high-performance in areas electric vehicle, aeronautics more-electric ship. This paper mainly demonstrates development this new drive system, introduces control method multi-phase motor home abroad, analyzes classic as well advanced rules decoupling control. To perfectly utilize we also conclude deficiency existed research put forward some technical outlook new-type structure.
Ultraviolet imaging technology is an effective method to detect the discharge of high-voltage electrical equipment. At present, photon number main characterize severity UV imaging. However, there a complex nonlinear relationship between parameter size and gain observation distance imager. The was quantified. In order quantify analysis discharge, based on OpenCV image processing technology, segmented by combined threshold method, region extracted multi-region contour algorithm; Using...