A5079581616- Advanced Neural Network Applications
- Video Surveillance and Tracking Methods
- Fire Detection and Safety Systems
- Calibration and Measurement Techniques
- Ga2O3 and related materials
- ZnO doping and properties
- Thermography and Photoacoustic Techniques
- Advanced Optical Sensing Technologies
- Autonomous Vehicle Technology and Safety
- GaN-based semiconductor devices and materials
- Industrial Vision Systems and Defect Detection
- Advanced Sensor Technologies Research
Shaoyang University
2018-2023
Central South University
2018-2022
In nighttime driving scenes, due to insufficient and uneven lighting, the scarcity of high-quality datasets, miss rate pedestrian detection (PD) is much higher than that daytime. Vision-based distance (DD) has advantages low cost good interpretability, but existing methods have precision, poor robustness, DD mostly performed independently PD. A narrowband near-infrared (NIR) camera an NIR lamp relieve impact noisy visible light (VIS) improve imaging quality been selected; a LiDAR obtain...
Pedestrians, motorist, and cyclist remain the victims of poor vision negligence human drivers, especially in night. Millions people die or sustain physical injury yearly as a result traffic accidents. Detection recognition road markings play vital role many applications such surveillance autonomous driving. In this study, we have trained nighttime road-marking detection model using NIR camera images. We modified VGG-16 base network state-of-the-art faster R-CNN algorithm by multilayer...
Ultraviolet detectors can be used in ultraviolet disinfection, missile guidance and short-wave communication fields. Here we have grown GaN (002) film by chemical vapor deposition (CVD), through depositing 100 nm ZnO as buffer layers on sapphire substrates magnetron sputtering. Using Ni the electrode, metal-semiconductor-metal (MSM) (UV) detector was prepared. The device dark current, photocurrent UV on/off current ratio were 5.19×10-9A, 2.52×10-7A 54, under 2V bias. Then, using Au...