A5047770388- Advanced Image and Video Retrieval Techniques
- Infrared Target Detection Methodologies
- Soil Moisture and Remote Sensing
- Electromagnetic Simulation and Numerical Methods
- High voltage insulation and dielectric phenomena
- Electromagnetic Scattering and Analysis
- Advanced SAR Imaging Techniques
- Antenna Design and Optimization
- Advanced Neural Network Applications
- Remote Sensing and Land Use
- Advanced Optical Sensing Technologies
- Microwave Engineering and Waveguides
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Antenna Design and Analysis
- Geophysical Methods and Applications
- Image Enhancement Techniques
- Optical Systems and Laser Technology
Shanghai Jiao Tong University
2020-2024
Shanghai Technical Institute of Electronics & Information
2021
The inspection of insulators and their defects is great significance for ensuring the safety stability power system. Small sample one main issues insulator defect detection based on neural network. In this research, we release a dataset self-explosive detection, provide benchmark improved YOLOv5, named Foggy Insulator Network (FINet). work, synthetic fog algorithm implemented optimized. An (SFID) with 13000 images constructed released. YOLOv5 network into SE-YOLOv5 by introducing channel...
Deep learning has significantly enhanced the performance of automatic target recognition (ATR) in synthetic aperture radar (SAR). However, concept model overinterpretation, characterized by classifiers discerning strong class evidence within image regions that lack semantically salient features related to (e.g., background clutter SAR images), undermined confidence reliability deep models. Previous studies predominantly relied solely on one interpretability method qualitatively identify key...
A novel surrogate model is proposed based on space mapping to maintain accurate predictions of the radar cross section over a relative large range frequencies. With sparse-sample full-wave solver and dense-sample ray-tracing solver, can be established with small computational burden. function introduced represent relationship between two such solvers overcome limitations conventional fixed linear form. Two numerical examples are used verify effectiveness method, which demonstrate its ability...
Spaceborne synthetic aperture radar (SAR) is an active microwave imaging equipment and has play a key role in many fields such as target dynamic monitoring 3D mapping, due to its unique advantage of observing all time all-weather. In this paper, structure design thermal deformation analysis large spaceborne millimeter wave phased array SAR antenna prescribed. The total heat per unit generated by the antenna, whose size about 590λ<inf>0</inf>×56λ<inf>0</inf>...
A data fusion architecture is proposed for the and speed tagging of moving object in both camera images two-dimensional radar detection results. After calibration between coordinate parameters, a Faster-Region Convolutional Neural Network (Faster-RCNN) used to determine point object, while simplified Recurrent (RNN) extract target trajectory. Experiment results show that method able tag range targets fast accurate manner.