A5016490930- Advanced Neural Network Applications
- Video Surveillance and Tracking Methods
- Modular Robots and Swarm Intelligence
- Image Enhancement Techniques
- Visual Attention and Saliency Detection
- Robotic Path Planning Algorithms
- Traffic Prediction and Management Techniques
- Advanced Image and Video Retrieval Techniques
- Random lasers and scattering media
- Neural Networks and Reservoir Computing
- Electrowetting and Microfluidic Technologies
- Metallurgy and Material Science
- Advancements in PLL and VCO Technologies
- Low-power high-performance VLSI design
- VLSI and Analog Circuit Testing
- Mobile Agent-Based Network Management
- Image and Object Detection Techniques
- Target Tracking and Data Fusion in Sensor Networks
- DNA and Biological Computing
- Gait Recognition and Analysis
- E-commerce and Technology Innovations
- Robotic Locomotion and Control
- Power Line Inspection Robots
- Smart Parking Systems Research
- Speech Recognition and Synthesis
Xi’an University of Posts and Telecommunications
2020-2025
Zhengzhou University of Light Industry
2020-2025
Zhejiang University
2019-2025
Zhongkai University of Agriculture and Engineering
2025
University of Illinois Urbana-Champaign
2025
North China Electric Power University
2018-2024
Institute of Microelectronics
2018-2023
Chinese Academy of Sciences
2018-2023
North China University of Technology
2021
Kunming University of Science and Technology
2021
Because insulators provide electrical insulation and mechanical support for electric transmission lines, these components are of paramount importance to safe reliable operations power systems. However, often considered be prone different faults, e.g., bunch-drop, which demands a novel solution accurate fault detection location. Current research efforts have primarily focused on the bunch-drop glass insulators, study ceramic has not been reported date. To this end, paper proposes an...
In practical applications of droplet actuation using digital microfluidic (DMF) systems based on electrowetting-on-dielectric (EWOD), various electrode failures can still arise due to diverse operational conditions. To improve transport efficiency, this study proposes a heuristic-elite genetic algorithm (HEGA) for path planning. We introduce heuristic method and bidirectional elite fragment recombination address the challenge poor initialization quality in algorithms, particularly complex...
In daily research engineers and technicians often face problems such as insufficient network equipment, difficult to deploy scenarios, test operation results. order solve these problems, better master the design method deployment process of integrated deepen understanding environment, equipment protocol interaction, combined with author’s own practical experience in engineering, this paper analyzes scenario requirements an enterprise, skillfully integrates virtual local area network, trunk,...
Abstract Chiral imaging, which utilizes chiral nanostructures to produce grayscale images, has attracted significant attention due its potential for precise object and environmental perception. However, conventional imaging techniques primarily rely on external light sources other necessary optical components create desired patterns, are compatible with device integration miniaturization. In this study, a passive technique is presented by fully exploiting the spontaneous thermal emission of...
In order to explore the application of image recognition model based on multi-stage convolutional neural network (MS-CNN) in deep learning intelligent commodity images and performance method, study, features color, shape, texture are first analyzed, basic structure (CNN) is analyzed. Then, 50,000 pictures containing different commodities constructed verify effect model. Finally, MS-CNN taken as research object for improvement influence label errors (p = 0.03, 0.05, 0.07, 0.09, 0.12) with...
Using an unmanned aerial vehicle (UAV) to collect data from wireless sensor networks deployed in the field, one of key tasks is plan path for collection so as minimize energy consumption UAV. At present, most existing methods generally take shortest flight distance optimal objective path. They simply believe that means least UAV and ignore fact changing direction (heading) can also consume UAV’s its flight. If be planned based on closer real situation, really reduced working efficiency...
A neural network model based on deep learning is utilized to explore the traffic sign recognition (TSR) and expand application of intelligent technology in field virtual reality (VR) image recognition, thereby assessing road safety risks promoting construction transportation networks. First, a dual-path CNN (TDCNN) TSR built convolutional (CNN), cost function accuracy are selected as indicators analyze training results model. Second, recurrent (RNN) long-short-term memory (LSTM) RNN assess...
A series of computer vision-based fire detection algorithms is proposed in this paper. These can be used parallel with conventional systems to reduce false alarms. They also deployed as a stand- alone system detect combined video acquisition device. This paper chiefly discussed the segmentation problem. It adopted L*a*b*, YCbCr color space and k- means clustering algorithm isolate forest detection. characterized by high stabilization versatility, improves real time accurate, has been fires...
This paper presents an innovative mechanical design of a reconfigurable miniature throwable robot. The robot keeps in spherical shape to buffer shock and protect inner sub-systems falling, it transforms two-wheeled mobile surveillance via wireless control. Two generations the have been constructed, details design, principle transformation servo-system for adjusting view level micro-camera are discussed. Experiments also presented, effectiveness has approved.
In order to solve the problem of USV dynamic path planning with towed safety boundary in complex marine environment, an adaptive algorithm is proposed. Firstly, obstacle avoidance and bounded tension cable array are analyzed respectively, then correspondent model condition established. Secondly, a fuzzy mechanism strategy developed automatically choose appropriate local searching radius according degree obstacles. Moreover, for some constraints such as large-angle turn demand process...
Compressive imaging that reduces the pixel numbers can lower cost and power consumption. Here, we use a random-PSF metasurface computational back-end to achieve compression rate of 50% for general scenes.
Functional engineering change order (ECO) has been an essential process in modern complex integrated circuit design. Finding a high-quality patch efficiently long challenge. This paper proposes FEEP, automatic and efficient synthesis-based functional ECO method. Structural pruning stratified searching techniques are proposed to minimize search space without extra logical equivalence checks. Moreover, we propose machine-learning-based two-stage size predictor that assists predicting quality....
This article endeavors to enhance image recognition technology within the context of Internet Things (IoT). A dynamic target detection training model is established through convolutional neural network (CNN) algorithm framework deep learning (DL). Three distinct configurations are proposed: a nine-layer convolution model, seven-layer and residual module model. Subsequently, simulation CNN based on optical imaging constructed, experiments conducted in scenarios simple salient environments,...
There are some methods such as mutual information, alignment metric and so on. But they aim at registration of multi-modal images whose sizes same. And there few researches on different. In this paper, an automatic method for based is proposed. order to look the best location two images, genetic algorithm used in iterative search. Alignment fitness function algorithm. It proved that different can be realized. it has high accuracy robustness.
Mini‐LED has developed rapidly by virtue of its advantages in display industry, including low power consumption, stability, high resolution and effect. In this paper, a video system based on FPGA with multiple mini‐LED displays is implemented. The builds the circuit structure multi‐screen splicing realizes flexible displays, so that can achieve effect expansion size. Multiple images stored SD card are read segmented into parts corresponding to number screens then output spliced screen for...
It is important to synchronize the data acquisition when digitizing PET event pulses with multiple detectors. In this paper, we design and implement a basic clock distribution module, module can provide 12 synchronized reset outputs. By connecting several modules, it practical build larger system in tree structure. The structure applicable digital of different scale, height 3, drive 1728 We built simplified three-level system, period standard deviations were measured be below 20 ps, skews...
The work aims to study the application of Deoxyribonucleic Acid (DNA) multi-source data storage in Digital Twins (DT) . Through investigation research status DT and DNA computing, puts forward concept for DT. Raptor code is improved from design direction degree distribution function, six function schemes are proposed turn process describing method. Additionally, a quaternary dynamic Huffman coding method applied storage, combined with concatenated as error correction code. Considering...