A5100749763- Advanced Algorithms and Applications
- Simulation and Modeling Applications
- Advanced Computational Techniques and Applications
- Hand Gesture Recognition Systems
- Industrial Technology and Control Systems
- Advanced Measurement and Detection Methods
- Advanced Control Systems Optimization
- Human Pose and Action Recognition
- Iterative Learning Control Systems
- Advanced Sensor and Control Systems
- Advanced Decision-Making Techniques
- Embedded Systems and FPGA Design
- Robotics and Sensor-Based Localization
- Evaluation and Optimization Models
- Domain Adaptation and Few-Shot Learning
- Radiomics and Machine Learning in Medical Imaging
- Power Systems and Technologies
- Multimodal Machine Learning Applications
- Gait Recognition and Analysis
- Industrial Automation and Control Systems
- Smart Agriculture and AI
- Hydraulic and Pneumatic Systems
- Environmental DNA in Biodiversity Studies
- Digital Radiography and Breast Imaging
- Vehicular Ad Hoc Networks (VANETs)
China University of Geosciences
2020-2025
Griffith University
2025
Guizhou Water Conservancy and Hydropower Survey and Design Institute
2024
Sun Yat-sen University
2022-2023
China Automotive Technology and Research Center
2023
Nanjing Agricultural University
2021
Shandong Institute of Automation
2020-2021
Qingdao Academy of Intelligent Industries
2020
Wuhan Ship Development & Design Institute
2020
Guizhou Minzu University
2019
6DoF object pose estimation is a foundation for many important applications, such as robotic grasping, automatic driving, and so on. However, it very challenging to estimate of transparent which commonly seen in our daily life, because the optical characteristics material lead significant depth error results false estimation. To solve this problem, two-stage approach proposed from single RGB-D image. In first stage, influence eliminated by segmentation, surface normal recovering, RANSAC...
Hand gesture recognition and hand pose estimation are two closely correlated tasks. In this paper, we propose a deep-learning based approach which jointly learns an intermediate level shared feature for these tasks, so that the task can be benefited from task. training process, semi-supervised scheme is designed to solve problem of lacking proper annotation. Our detects foreground hand, recognizes gesture, estimates corresponding 3D simultaneously. To evaluate performance state-of-the-arts,...
A new model based nonlinear iterative learning controller to replicate multiple reference trajectory is presented. The approach consists of a constrained Gauss-Newton numerical optimization algorithm, which able inverse system accurately. It uses the moving horizon idea from MPC improve computational efficiency. Combined with class state-space and an existing control scheme, this allows accurate tracking system. Numerical results are used illustrate algorithm.
<abstract> <b><sc>Abstract.</sc></b> It is a challenging task to achieve accurately fruit recognition in 3D space for the robotic picking. This paper proposed an automatic segmentation and method of apple picking from point cloud data. First, HSI based region growing was segment data tree. Next, improved descriptor with fusion color features geometry computed each segmented clouds. And then, support vector machine optimized by genetic algorithm (GA-SVM) classifier constructed recognize...
Intertidal creek ecosystems in saltmarsh wetlands provide crucial ecological services, with burrows facilitating material exchange between these environments. Investigating burrow distribution tidal creeks is essential, yet the underlying mechanisms remain poorly understood. Our study conducted field observations during two seasons, dividing typical section into microhabitats: flat (F), gentle edge (GE), slope (GS), steep (SS), and (SE). Comprehensive analysis of sediment samples encompassed...
The plant communities and wild animals in the Gobi desert outside oases west of Hexi Corridor are an indispensable important part human living environment. Affected by global warming, spatial temporal distribution patterns precipitation have changed significantly. This article uses daily data from 6 hydro-logical stations middle lower reaches Shule River Basin 1957 to 2018, 3 meteorological 1980 2018. Through research on annual scale anomaly percentage, effective days distribution, drought...
Breast ultrasound is recommended for early breast cancer detection in China, but the rapid increase imaging data burdens sonographers. This study evaluated agreement between artificial intelligence (AI) software and sonographers analyzing nodule features. images from two hospitals Shanghai were analyzed by both features including echotexture, echo pattern, orientation, shape, margin, calcification, posterior attenuation. Agreement was compared using proportion of Kappa, with analysis time...
In the context of intelligent vehicle lane changing, a trajectory with superior performance is typically planned in accordance constraints change path. This allows to execute along preplanned trajectory. Model Predictive Control (MPC) stands out among other control methods due its ability incorporate variety into process. ensures that model aligns actual situation, rendering results practically significant. Consequently, this study employs MPC method design path tracking controller for...
With the advent of internet things, intelligent robots and automatic guided vehicles (AGVs) have developed rapidly. This paper proposes a landmark-based visual positioning system for an AGV used transporting goods in large warehouses. The runs along route that consists artificial landmarks, with camera capturing images ground. Every time passes through landmark, position orientation deviations between landmark can be calculated based on local features, including edges corners. To maintain...
Science and technology can play a positive role in agricultural production, the most representative of which is new intelligent system greenhouses. The greenhouse uses automatic control such as Internet Things big data to optimize crop growth environment. This article designed set an system, automatically detects light intensity, soil humidity, air temperature, carbon dioxide value, etc., through single-chip processing, adjust execution unit so that parameters are always kept under...
Electric tractors (ETs) have attached significant attention in recent years due to the increasingly stringent exhaust emission policies all over world. However, testing and evaluating possible architectures component sizes for ETs would be a time-consuming expensive task. Establishing comprehensive test bench that can implement different powertrain as much is thus an important challenging In this study, novel reconfigurable platform proposed according analysis results of topologies ETs....
The training of depth image-based hand pose estimation model typically relies on real-life datasets which are expected to be 1) largescale and cover a diverse range poses shapes, 2) always come with high-precision annotations. However, existing in reality rather limited the above regards due multitude practical constraints, time cost being major concerns. This observation motivates us propose an alternative approach, where is primarily trained synthesized images that closely mimicking...
Dimension feature recognition plays an important role in 2D engineering drawings. At present, most dimension methods, the first step is of whole feature, and then to extract information. These methods have some disadvantages, especially exploded recognition. Aimed solve this problem, paper introduces a new algorithm intelligent According position relationship with basic elements dimension, recognized bottom-up approach. This method can solves problem recognizing
This paper analyzes the major influencing factors of traffic safety on basis present safety, and establishes a evaluation system. Especially, existed methods, this puts forward new system —a method called integrated road that based Delphi-AHP-Fuzzy. When using method, combines Delphis, Analytic Hierarch Process (AHP), Fuzzy Comprehensive Evaluating Matrix in an way. In end paper, typical example is used to demonstrate actual procedure effective, simple intuitive have no dependence historical...
A kind of satellite navigation transmission system based on DSP and Ethernet hardwired TCP/IP stack chip W5100 is designed to satisfy the long distance abundant data required in signal processing. The made up central processor a number processing units. Each unit collects processes information under control DSP, sends with flexible fast ability W5100. simple efficient, speed every can reach 3Mbytes/s, which satisfies design require. has been put practical application.