A5000926804- Robotics and Sensor-Based Localization
- Robotic Path Planning Algorithms
- Welding Techniques and Residual Stresses
- Industrial Vision Systems and Defect Detection
- Surface Roughness and Optical Measurements
- Image Processing Techniques and Applications
- Advanced Image Processing Techniques
- Advanced Image and Video Retrieval Techniques
- Ultra-Wideband Communications Technology
- Indoor and Outdoor Localization Technologies
- 3D Surveying and Cultural Heritage
- Microwave Imaging and Scattering Analysis
- Advanced Image Fusion Techniques
- Microfluidic and Capillary Electrophoresis Applications
- Target Tracking and Data Fusion in Sensor Networks
- Photoacoustic and Ultrasonic Imaging
- Advanced Neural Network Applications
- Microfluidic and Bio-sensing Technologies
- Hydrogen embrittlement and corrosion behaviors in metals
- Non-Destructive Testing Techniques
- Speech and Audio Processing
- Remote Sensing and LiDAR Applications
- Electrical and Bioimpedance Tomography
- GNSS positioning and interference
- Muscle activation and electromyography studies
Central South University
2023-2024
Key Laboratory for High Strength Lightweight Metallic Materials of Shandong Province
2023-2024
Southeast University
2022
University of Roehampton
2022
This work aims to improve the low accuracy caused by interference information during real-time weld surface detection. Firstly, a dataset with 7580 pictures containing 8 types of defects was established. Secondly, an improved detection model named Yolo-MSAPF designed based on Yolo-v5 and verified self-established database. Finally, system built analyze performance in industrial environment. The design principle is eliminate but enhance necessary features each scale as much possible...
Abstract The essence of the difficulties for weld surface detection is that there a lot interference information during detection. This study aims to enhance accuracy while keeping great deployment capabilities model defects. To achieve this goal, an improved Yolo-GCH proposed based on stable and fast Yolo-v5. improvements primarily involve introducing graph convolution network combined with self-attention mechanism in head part (i.e., GCH). component focuses improving insufficient...
In this letter, we propose a novel Non-Line-of-Sight (NLOS) identification and error-mitigation method for dynamic object positioning ultra-wideband (UWB) ranging. By utilizing inverse estimation on known Anchor Points (APs) improved robust unscented Kalman filter (IRUKF), while fusing Gyroscope Accelerometer data, the proposed technology identifies compensates NLOS occlusions between tag APs, reducing errors. The approach has been verified through simulation experiment, with precision of...
Ultra-Wideband (UWB) technology is crucial for indoor localization systems due to its high accuracy and robustness in multipath environments. However, Non-Line-of-Sight (NLoS) conditions can cause UWB signal distortion, significantly reducing positioning accuracy. Thus, distinguishing between NLoS LoS scenarios mitigating errors enhancing system performance. This research proposes a novel 1D-ConvLSTM-Attention network (1D-CLANet) extracting temporal channel impulse response (CIR) features...
Bioinformation is information generated from biological movement. By using a variety of modern technologies, we can use this to form meaningful model for researchers study. An electromyographic (EMG) signal one type bioinformation that used in many areas help people study human muscle This both clinical and industrial areas. EMG very complicated signal, so processing it vital. The signals divided into collection, denoising, decomposition, feature extraction classification steps. In article,...
This paper proposes a novel path planner (i.e., E-Planner) based on the visibility graph for car-like mobile robots operating in unknown environments. To improve efficiency of building graph, an obstacle contour optimization approach is designed to reduce redundant vertices and edges. Moreover, breadth-first search combined with prioritized exploration timer mechanism employed decrease waste space path-searching process. Additionally, pure tracking method dynamic sight distance proposed...
In this study, we developed an inspection system integrating multiple modes for the detection of complex weld surface defects. On one hand, integrates a total 4 including offline and online detection. other can display visualization quantitative results classification position defects in real-time, make overall evaluation based on these results. It provides detailed suggestions new ideas high integration intelligence defect system.
This paper presents a method to improve the sensitivity of an impedance flow cytometry (IFC) chip by tuning its parasitic capacitance and operating frequency. First, equivalent model IFC was used study effect, numerical simulation predict frequency shift various electrode arrangements; Secondly, we validated our theoretical using three types arrangements fabricated soft lithography techniques; Thirdly, characterized responses device. By staggered configuration, reduced 57.7%, while center...