A5100641501- Advanced Vision and Imaging
- Industrial Vision Systems and Defect Detection
- Advanced Sensor and Control Systems
- 3D Shape Modeling and Analysis
- Soil Carbon and Nitrogen Dynamics
- Semiconductor Lasers and Optical Devices
- Advanced Neural Network Applications
- Medical Image Segmentation Techniques
- Robotic Path Planning Algorithms
- Remote-Sensing Image Classification
- Image Retrieval and Classification Techniques
- Engineering and Test Systems
- Advanced Computational Techniques and Applications
- Generative Adversarial Networks and Image Synthesis
- Advanced Optical Sensing Technologies
- Soil erosion and sediment transport
- Image Enhancement Techniques
- Date Palm Research Studies
- Advanced Image Fusion Techniques
- Smart Agriculture and AI
- Millimeter-Wave Propagation and Modeling
- Brain Tumor Detection and Classification
- Robot Manipulation and Learning
- Enhanced Oil Recovery Techniques
- Advancements in PLL and VCO Technologies
Hongzhiwei Technology (China)
2024
Peking University
2023-2024
Northwestern Polytechnical University
2023
Huawei Technologies (United Kingdom)
2023
University of Science and Technology of China
2022
Chinese University of Hong Kong
2020-2022
China Academy of Space Technology
2021
Wuhan University of Science and Technology
2020
University of Oxford
2020
Jiangsu University
2020
Forestation is regarded as an effective strategy for increasing terrestrial carbon sequestration. However, its sink potential remains uncertain due to the scarcity of large-scale sampling data and limited knowledge linkage between plant soil C dynamics. Here, we conduct a survey 163 control plots 614 forested involving 25304 trees 11700 samples in northern China fill this gap. We find that forestation contributes significant (913.19 ± 47.58 Tg C), 74% which stored biomass 26% organic carbon....
Enhancing terrestrial carbon (C) stock through ecological restoration, one of the prominent approaches for natural climate solutions, is conventionally considered to be achieved an pathway, i.e., increased plant C uptake. By conducting a comprehensive regional survey 4279 1 × m 2 plots at 517 sites across China’s drylands and 13-y manipulative experiment in semiarid grassland within same region, we show that greater soil ecosystem stocks restored than degraded lands result predominantly from...
Medical image registration is a fundamental task in medical analysis, aiming to establish spatial correspondences between paired images. However, existing unsupervised deformable methods rely solely on intensity-based similarity metrics, lacking explicit anatomical knowledge, which limits their accuracy and robustness. Vision foundation models, such as the Segment Anything Model (SAM), can generate high-quality segmentation masks that provide structure addressing limitations of traditional...
Building change detection has always been an important research focus in production and urbanization. In recent years, deep learning methods have demonstrated a powerful ability the field of detecting remote sensing changes. However, due to heterogeneity characteristics buildings, current do not present effective means perceive building changes or fuse multi-temporal features, which leads fragmented incomplete results. this article, we propose multi-branched network structure semantic...
Object-detection methods based on deep learning play an important role in achieving machine automation. In order to achieve fast and accurate autonomous detection of stacked electronic components, instance segmentation method improved Mask R-CNN algorithm was proposed. By optimizing the feature extraction network, performance improved. A dataset components containing 1200 images (992 × 744 pixels) developed, four types were included. Experiments showed model superior speed while being more...
Abstract We study a general form of gears known as non‐circular that can transfer periodic motion with variable speed through their irregular shapes and eccentric rotation centers. To design functional is nontrivial, since the gear pair must have compatible shape to keep in contact during motion, so driver push follower rotate via bounded torque motor exert. address challenge, we model geometry, kinematics, dynamics gears, formulate problem optimization, identify necessary independent...
This paper investigates a multiuser millimeter-wave (mmWave) uplink system in which each user is equipped with multi-antenna fluid antenna (FAS) while the base station (BS) has multiple fixed-position antennas. Our primary objective to maximize capacity by optimizing transmit covariance matrices and position vectors of users jointly. To gain deeper insights, we commence deriving upper bounds approximations for maximum capacity. Then delve into maximization problem. Beginning simple scenario...
AbstractThe production of hydrogen sulfide (H2S) had caused huge economic losses and security risks to oil fields, which attracted the attention many researchers. Studies focused on hypersaline reservoirs were scarce. Here, we simulated H2S inhibition by adding NO3- NO2- in anaerobic bottles up-flow sand-packed bioreactor laboratory using water samples from injection wells Jianghan reservoir. The microbial characteristics, pH, salinity, sulfate, sulfide, measured at every stage. content...
We introduce the first neural optimization framework to solve a classical instance of tiling problem. Namely, we seek non-periodic an arbitrary 2D shape using one or more types tiles: tiles maximally fill shape's interior without overlaps holes. To start, reformulate as graph problem by modeling candidate tile locations in target nodes and connectivity between edges. Further, build convolutional network, coined TilinGNN, progressively propagate aggregate features over edges predict...
Abstract We study structural rigidity for assemblies with mechanical joints. Existing methods identify whether an assembly is structurally rigid by assuming parts are perfectly rigid. Yet, identified as may not be that “rigid” in practice, and existing cannot quantify how is. address this limitation developing a new measure, worst‐case , to the of largest possible deformation undergoes arbitrary external loads fixed magnitude. Computing non‐trivial due non‐rigid different joint types. thus...
In automatic welding processing, high temperature likely induces local deformation of the weldment, resulting in accuracy deviation and quality. Therefore, research on weld tracking technology is significant. Most current trackers are installed at end torch. Due to tracker volume, interference often occurs operations under complex conditions. This paper designs develops a seam system for human-computer interactive mobile robots. The spatial coordinates were obtained by camera calibration,...
Recognizing various surgical tools, actions and phases from surgery videos is an important problem in computer vision with exciting clinical applications. Existing deep-learning-based methods for this either process each video as a series of independent images without considering their dependence, or rely on complicated deep learning models to count dependence frames. In study, we revealed exploratory data analysis that enjoy relatively simple semantic structure, where the presence tools can...
This paper presents a computational pipeline for creating personalized, physical LEGO ®1 figurines from user-input portrait photos. The generated figurine is an assembly of coherently-connected ® bricks detailed with uv-printed decals, capturing prominent features such as hairstyle, clothing style, and garment color, also intricate details logos, text, patterns. task non-trivial, due to the substantial domain gap between unconstrained user photos stylistically-consistent models. To ensure...
In order to solve the problems of different visual angles, such as large brightness and colour difference, obvious image splicing trace, an algorithm ring fusion based on correction is proposed in this paper. According paper, a Matlab toolbox was made for simulation experiment. Structural similarity (SSIM) used evaluate effectiveness method reducing illumination contrast difference mean value SSIM 0.8166. The measured development board accelerated by graphic processing unit achieve computing...
Data collection in the important position of modern industrial production and scientific research is increasingly outstanding, as well real-time collection, transmission processing high speed data acquisition requirements are constantly improve. In addition, for different occasions, system sampling parameters different. PLD a universal integrated circuit, its logic function according to user's programming on device, this paper introduces high-speed based PLD, mainly expounds software design...
Aiming at the problem that: when stirring blades of a vertical mixer stop, spatial position is random and surface configuration complex. Robots cannot clean blade surfaces effectively. In this paper, camera installed end robot. Using robot visual servo, successfully reads designed rotation marker code; obtains exact posture from code. Finally, Retrieve corresponding cleaning program to implement robot's cleaning. end, experimental platform built prove that contact force between normal during...
The rapid development of 3D imaging technology has made the representation objects through point clouds and associated triangulations a routine process. Data on human faces is common example. analysis datasets involving multiple individuals requires models for inherent variability in shape, both overall image sub-features interest. In this paper, we propose an algorithm extraction features dissection along particular curve. This mesh-reconstruction creates component parts whose edges are...