A5054363909- Phosphorus and nutrient management
- Infrastructure Maintenance and Monitoring
- Bauxite Residue and Utilization
- Image and Video Quality Assessment
- Advanced Vision and Imaging
- Advanced Image Processing Techniques
- Coal and Coke Industries Research
- Iron and Steelmaking Processes
- Video Coding and Compression Technologies
- GaN-based semiconductor devices and materials
- Wireless Communication Networks Research
- Video Surveillance and Tracking Methods
- Advanced Algorithms and Applications
- Advanced Sensor and Control Systems
- Image Enhancement Techniques
- Speech and Audio Processing
- Advanced Computational Techniques and Applications
- Radio Frequency Integrated Circuit Design
- Image and Signal Denoising Methods
- Geoscience and Mining Technology
- Smart Grid and Power Systems
- Industrial Technology and Control Systems
- Advanced Neural Network Applications
- Vehicle License Plate Recognition
- Advanced Image Fusion Techniques
Qinghai University
2022-2024
Hunan University
2024
Qinghai University for Nationalities
2013-2023
Jinan University
2022
Sinopec (China)
2005-2011
China Design Group (China)
2002-2004
Nanjing Institute of Industry Technology
2001
Restoring underwater image from a single is know to be ill-posed, and some assumptions made in previous methods are not suitable for many situations. In this paper, we propose method based on blue-green channels dehazing red channel correction restoration. Firstly, recovered via algorithm an extension modification of Dark Channel Prior algorithm. Then, corrected following the Gray-World assumption theory. Finally, order resolve problem which regions may look too dim or bright, adaptive...
Restoring underwater image from a single is known to be an ill-posed problem. Some assumptions made in previous methods are not suitable many situations. In this paper, effective method proposed restore images. Using the quad-tree subdivision and graph-based segmentation, global background light can robustly estimated. The medium transmission map estimated based on minimum information loss principle optical properties of imaging. Qualitative experiments show that our results characterized by...
In this article, we present the analysis, design, and implementation of a wideband 10-W monolithic microwave integrated circuit power amplifier (PA), fabricated in low-cost 0.1-μm gallium nitride (GaN) on Si technology. The design is focused realization low-loss impedance transformation networks across 2-20 GHz using reactive matching (RM) technique. two-stage GaN PA achieves an average output 40.1 dBm peak 41.6 at 13 GHz, CW-mode operation, with smallsignal gain S <sub...
The design and implementation of a high-gain high-power, efficient broadband monolithic microwave integrated circuit (MMIC) power amplifier (PA) using 0.1-μm GaN HEMT process is presented. For the first time, triple-stacked transistors used in non-uniform distributed to achieve high efficiency simultaneously. Further, similar stacked cell with feedback technique adopted as compact driver enhance gain performance. Measured results MMIC PA across 2 18 GHz band show at least 29 ± 1 dB...
Since the characteristics of wireless channel are open and broadcasting, networks very vulnerable to be attacked via eavesdropping, jamming, interference. As traditional secure technologies not suitable for PHY layer networks, physical-layer security issues become a focus attention. In this paper, we firstly identify summarize threats vulnerabilities in networks. Then, give holistic overview schemes, which divided into three categories: spatial domain-based, time frequency domain-based....
This paper presents the design and measured results of a monolithic microwave integrated circuit (MMIC) T/R front-end module (FEM) suitable for S band to Ku applications, using 0.1µm GaN HEMT process. For first time, successfully 3-stacked non-uniform distributed power amplifier (SNDPA), two-stage 2-stacked low noise (LNA) switch in one MMIC, obtain ultra-broadband response Tx mode low-noise low-consumption performance Rx mode, simultaneously. Measured MMIC FEM across 2 18 GHz show that...
Two new citrinin dimers bearing a 6,6-spiroketal moiety (1 and 2) four known analogues (3–6), together with 18 monomers (7–24), were isolated from the culture of hydrothermal vent-associated fungus Penicillium citrinum Y34. Their structures identified by extensive spectroscopic analyses, 13 C NMR calculation in combination DP4+, linear correlation coefficient ( R 2 ), mean absolute error (MAE) values electronic circular dichroism (ECD) calculation. The α -glucosidase ATP-citrate lyase (ACL)...
The network coding (NC) could improve the throughput and reliability. However, it is vulnerable for pollution attack eavesdropping attack. What's more, due to high complexity of traditional secure (SNC) schemes, many them are not suitable resource-constrained wireless sensor (WSN). In this paper, a low based on double primes (DP-SNC) proposed defense against DP-SNC, DP-sequence first constructed by operation randomized transform, then encoding process divided into two phases: source node...
By combining the finite element analysis (FEA) and artificial neural network (ANN) technique, complete prediction of interconnect reliability for a monolithic microwave integrated circuit (MMIC) power amplifier (PA) at both direct current (DC) alternating (AC) operation conditions is achieved effectively in this article. As example, MMIC PA modelled to study electromigration failure interconnect. This first time an DC AC simultaneously. training data from FEA, high accuracy ANN model...
Image segmentation of sea-land remote sensing images is great importance for downstream applications including shoreline extraction, the monitoring near-shore marine environment, and target recognition. To mitigate large number parameters improve accuracy, we propose a new Squeeze-Depth-Wise UNet (SDW-UNet) deep learning model image segmentation. The proposed SDW-UNet leverages squeeze-excitation depth-wise separable convolution to construct modules, which enhance capacity in combining...
To strengthen data privacy protection, and improve mining accuracy, random response mode is adopted to design protection method based on association rules. Granular computing technology are applied fields of rules in mining, researched more extensive way from another perspective this paper. Firstly, partial concealing conceal transform original security; secondly, associated frequent item set utilized construct simple efficient algorithm; finally, algorithm proposed verified have higher...
Rock mass fractures are one of the main factors leading to slope instability, and detection rock can predict instability a certain extent. The rapid development deep learning has provided low-cost efficient method for fractures. This paper employs DenseNet121 model InceptionV2 fractures, improves models by incorporating an attention mechanism. dataset consists masses with from various regions enhance model's applicability in different scenarios. Experiments have revealed that family exhibits...
The quality of monitoring data sources has an important impact on road recognition and human vehicle detection performance in the field smart transportation. Aiming at problem large light disturbance low contrast urban images practical applications, this paper proposed a method for low-light image enhancement roads based Zero-DCE. First, constructed set, improved network model was designed idea no-reference depth curve estimation; then, hyperparameters are ablated, end-to-end training can be...
With the acceleration of intelligent transportation in new era, license plate recognition technology has become more and extensive. Although existing can show good performance for specific scenes traditional recognition, it poor robustness applicability complex external environments such as arbitrary shooting angles, different lighting conditions, picture quality. To solve above problems, this paper proposes an improved algorithm suitable various types plates. First, SSD model is used...