A5100984432- Fire Detection and Safety Systems
- Automated Road and Building Extraction
- Remote Sensing and LiDAR Applications
- Maritime Navigation and Safety
- Underwater Acoustics Research
- Air Quality Monitoring and Forecasting
- Advanced Image Fusion Techniques
- Image and Object Detection Techniques
- Advanced X-ray Imaging Techniques
- Advanced Image Processing Techniques
- Radiative Heat Transfer Studies
- Medical Image Segmentation Techniques
- Smart Agriculture and AI
- Video Surveillance and Tracking Methods
- Anomaly Detection Techniques and Applications
- ECG Monitoring and Analysis
- Non-Invasive Vital Sign Monitoring
- Image Processing Techniques and Applications
- Advanced Image and Video Retrieval Techniques
- Advanced machining processes and optimization
- Image and Signal Denoising Methods
- Image Enhancement Techniques
- Plant Disease Management Techniques
- Maritime and Coastal Archaeology
- Heart Rate Variability and Autonomic Control
China University of Petroleum, East China
2025
Institute of Computing Technology
2024
Shandong University of Science and Technology
2021
Henan Institute of Geological Survey
2017
PLA Information Engineering University
2016
Harbin University of Commerce
2015
Deep learning-based SAR oil spill detection faces significant challenges due to limited labeled training data. To address this, we propose SinGAN-Labeler, an enhanced framework that generates high-quality synthetic images and their labels from minimal input. The model integrates adaptive module automate scale parameter optimization, accelerating training, a hybrid attention combining spatial, channel, global contextual mechanisms enhance feature extraction. By leveraging multi-scale with...
In optical remote sensing images, the aircraft to be detected is very small; external environmental factors such as cloud occlusion, aircraft, and site background are easily fused; interference of objects has a great impact on characteristics in images. response above problems, we designed detection method based deep learning. First, ensure feature extraction capability limit number calculations network, LightNet v2 network unit designed, it constitutes an efficient backbone network....
Navigation landmark features such as docks are often used in ships for localization and searching shore targets during the voyage, which of great economic military significance. Continuous complete dock data could hardly be extracted by existing coast extraction method, because spatial relationships other characteristics ignored only considers grayscale remote sensing image .A method based on waterline perceptual organization is proposed this paper. To make full use relationships, introduced...
Aiming at the problems of color distortion, nonuniform illumination, and low contrast caused by degradation underwater images, an image enhancement method (MSFF-GAN) based on generative adversarial network was proposed. A multiscale featured fusion generator is designed, which improves ability to use different scale features model ensures that generated retains more detailed information. The residual dense module constructed solve problem characteristics extracted slower. In discriminator,...
After the characteristics of geodesic active contour model (GAC), Chan-Vese (CV) and local binary fitting (LBF) are analyzed, based on regions edges is combined with image segmentation method quad-tree, a waterline extraction quad-tree multiple proposed in this paper. Firstly , provides an initial according to segmentation; secondly, new signed pressure force (SPF) function global statistics information CV LBF has been defined, then, edge stopping function(ESF) replaced by SPF function,...
Weld seam recognition is the key technique in visual tracking research.Based on weld image features, preprocessing and linear feature extraction method of are designed.The extended adaptive median filtering modified Otsu threshold selection for segmentation used pre-processing process, to remove noise compress data.Through comparing with several other edge operators, Roberts operator chosen achieve goal detecting welding edge.Finally, information extracted through Hough transform a satisfied...