A5021092666- Remote-Sensing Image Classification
- Remote Sensing and Land Use
- Maritime Navigation and Safety
- Aerospace Engineering and Energy Systems
- Image and Signal Denoising Methods
- Advanced Image Fusion Techniques
- Iterative Learning Control Systems
- Advanced Algorithms and Applications
- Infrared Target Detection Methodologies
- Advanced Sensor and Control Systems
- Underwater Vehicles and Communication Systems
- Inertial Sensor and Navigation
- Industrial Technology and Control Systems
- Real-time simulation and control systems
- Face and Expression Recognition
- Advanced Surface Polishing Techniques
- Advanced machining processes and optimization
- Ship Hydrodynamics and Maneuverability
- Aerodynamics and Fluid Dynamics Research
- Advanced Chemical Sensor Technologies
- Heart Rate Variability and Autonomic Control
- Robotic Path Planning Algorithms
- Advanced Computational Techniques and Applications
- Advanced Control Systems Optimization
- Optical Polarization and Ellipsometry
Northwestern Polytechnic University
2025
China University of Petroleum, East China
2025
Northwestern Polytechnical University
2017-2024
Qingdao University of Science and Technology
2024
Guizhou Electric Power Design and Research Institute
2024
Shanghai Advanced Research Institute
2023
Chinese Academy of Sciences
2023
Shanghai University
2014-2023
Mongolian National University
2023
Space Engineering University
2022
Myocardial infarction is one of the most threatening cardiovascular diseases for human beings. With rapid development wearable devices and portable electrocardiogram (ECG) medical devices, it possible conceivable to detect monitor myocardial ECG signals in time. This paper proposed a multi-channel automatic classification algorithm combining 16-layer convolutional neural network (CNN) long-short term memory (LSTM) I-lead ECG. The preprocessed raw data first extract heartbeat segments; then...
This paper presents the first challenge on demosaicing of natural spectral images for snapshot hyperspectral imaging systems (HIS) which utilize a multi-spectral filer array (MSFA), i.e., recovery whole-scene information from spatially sub-sampled information. expands "ARAD_1K" data set to first-of-its-kind large-scale multispectral filter scenes containing 1,000 images. Challenge participants were required recover synthetically generated MSFA simulating capture by known calibrated mosaic...
Aiming at the problems of low accuracy strawberry fruit picking and large rate mispicking or missed picking, YOLOv5 combined with dark channel enhancement is proposed. In “Fengxiang” strawberry, criterion “bad fruit” added to conventional three criteria ripeness, near-ripeness, immaturity, because some bad fruits are close color ripe fruits, but small dry. The training four kinds strawberries different ripeness above 85%, testing 90%. Then, meet demand all-day address problem illumination...
This paper presents a mosaic convolution-attention network (MCAN) for demosaicing spectral images captured using multispectral filter array (MSFA) imaging sensors. MSFA-based systems acquire information of scene in single snap-shot operation. A complete image is reconstructed by an image. To avoid aliasing and artifacts demosaicing, we utilize joint spatial-spectral correlation raw The proposed MCAN includes convolution module (MCM) attention (MAM). MCM extracts features via learning...
This paper presents a deep learning-based spectral demosaicing technique trained in an unsupervised manner. Many existing techniques relying on supervised learning with synthetic images, often underperform real-world especially as the number of bands increases. comprehensive (USD) framework based characteristics mosaic images. encompasses training method, model structure, transformation strategy, and well-fitted selection strategy. To enable network to dynamically correlation while...
AbstractIn this paper parameter optimization through a quadratic performance index is introduced as method to establish new iterative learning control law. With algorithm, monotonic convergence of the error zero guaranteed if original system discrete-time LTI and it satisfies positivity condition. If not positive, two methods are derived make positive. The effect choice weighting parameters in on rate analysed. As result adaptive weights improve properties algorithm. A high-order version...
Laplacian support vector machine (LapSVM), as a benchmark method, which includes an additional regularization term with graph Laplacian, has been successfully applied to remote sensing image classification. However, using the Euclidean distance construct weights, in LapSVM may not really represent inherent distribution of data. In this paper, optimized LapSVMs are developed for semisupervised hyperspectral classification, by introducing metric learning instead traditional is used existing...
Global route planning has garnered global scholarly attention as a crucial technology for ensuring the safe navigation of intelligent ships. The comprehensive influence time-varying factors such water depth, prohibited areas, navigational tracks, and traffic separation scheme (TSS) on ship in coastal not been fully considered existing research, study method from perspective practical application is still needed. In this paper, based human-like thinking sailing scenarios proposed. Based...
With the rapid development of shipping industry, number ships is continuously increasing, and maritime accidents happen frequently. In recent years, computer vision drone flight control technology have developed, making drones widely used in related fields such as target detection. Compared to cameras fixed on ships, a greater flexibility wider field view provided by equipped drones. However, there are still some challenges high-altitude detection with Firstly, from top-down view, shapes...
Abstract In the process of ultra-deepwater drilling, safe density window is narrow, and there are issues with large temperature differences alternating high low temperatures, making wellbore pressure control challenging highly prone to complex lost circulation incidents. To achieve efficient accurate identification timely monitoring incidents essential. address current insufficient real-time performance accuracy in for this paper proposes a multi-parameter method based on support vector...
MOF-based sensors capable of effectively and stably detecting toxic species in water have attracted huge attention terms improving environmental monitoring levels quality. Combining the flexibility structure modifying pore surface, a multiwalled divider installation (MWDI) strategy is proposed used for property enhancement. We herein report three metal–organic frameworks (MOFs) 1–3 based on C3 symmetry organic phosphonic ligand with topology increased from 3,6-connected to 3,8-connected....
Local path planning, as an essential technology to ensure intelligent ships’ safe navigation, has attracted the attention of many scholars worldwide. In most existing studies, impact COLREGS received limited consideration, and there is insufficient exploration method in complex waters with multiple interfering ships static obstacles. Therefore, this paper, a generation for time–space overlapping equivalent obstacle line multi-ship encounter scenarios where both dynamic obstacles coexist...
The driver in road hypnosis has not only some external characteristics, but also internal characteristics. External features have obvious manifestations and can be directly observed. Internal do cannot They need to measured with specific instruments. Electroencephalography (EEG), as an feature of drivers, is the golden parameter for drivers’ life identification. EEG great significance identification hypnosis. An method based on human data proposed this paper. drivers collected through...
Recently, research in semisupervised learning (SSL) based on sparse representation has shown huge potential for many classification tasks. In this paper, we address a hyperspectral image by integrating L1-graph and SSL. We propose method with which more attractive merits than traditional graph method, such as parameter free, sparsity robustness. Our firstly obtains the weights solving L1 optimization problem, then generates way of SSL to deal images. The experiments are designed cope...
This paper presents a hyperspectral image (HSI) reconstruction technique based on physics-driven optimization of multispectral filter array (MSFA) patterns. The encoding HSIs using an MSFA and their decoding through deep learning has gained increasing attention. However, previous studies have seldom explored pattern from physical perspective during the process. In this paper, we apply spectral sensitivity function (SSF) response model to generate MSFA, goal encoder extends SSF structural...
Giant cell tumor of bone (GCTB) is potentially malignant with a high risk recurrence. Conventional GCTB diagnosis often empirical based on histopathology low efficiency. Reliable auxiliary system can improve the efficiency, but it need accurate detection. Changes micro-structure tissue alter scattering properties that lead to variation polarization state tissue. This paper proposed detection method by using Mueller matrix microscopic (MMPM) imaging and multi-parameters fusion network. First,...
To achieve self-healing in the smart grid, microgrid can be switched to islanded state for effective autonomy event of disasters or attacks. Based on data collected by meters, support an efficient energy scheduling strategy guide both suppliers and consumers so that electricity used efficiently. Significant research efforts have been carried out regarding scheduling. Nonetheless, majority existing strategies failed prioritize consumers. In fact, consumption order different users should...