A5070423020- Video Surveillance and Tracking Methods
- Advanced Neural Network Applications
- Advanced Image Processing Techniques
- Autonomous Vehicle Technology and Safety
- Advanced Image Fusion Techniques
- Fire Detection and Safety Systems
- Automated Road and Building Extraction
- Remote-Sensing Image Classification
- Image and Signal Denoising Methods
- Anomaly Detection Techniques and Applications
- Traffic Prediction and Management Techniques
- Remote Sensing and LiDAR Applications
- Vehicle Noise and Vibration Control
- Vehicle License Plate Recognition
- Advanced Vision and Imaging
- Civil and Geotechnical Engineering Research
- Image Processing Techniques and Applications
- Air Quality Monitoring and Forecasting
- Atmospheric and Environmental Gas Dynamics
- Smart Agriculture and AI
- Human Mobility and Location-Based Analysis
- Transport Systems and Technology
- Sustainable Industrial Ecology
- Greenhouse Technology and Climate Control
- Agriculture, Plant Science, Crop Management
Yantai University
2021-2025
Chongqing Jiaotong University
2009-2022
Shanghai University of Engineering Science
2021-2022
Chang'an University
2018-2020
Hunan Communications Research Institute
2010
Tongji University
2008
Remote sensing image segmentation plays an important role in many industrial-grade processing applications. However, the problem of uncertainty caused by intraclass heterogeneity and interclass blurring is prevalent high-resolution remote images. Moreover, complexity information images leads to a large amount background around objects. To solve this problem, new fuzzy convolutional neural network proposed paper. This resolves ambiguity feature introducing neighbourhood module deep learning...
Forest fires are a vulnerable and devastating disaster that pose major threat to human property life. Smoke is easier detect than flames due the vastness of wildland scene obscuring vegetation. However, shape wind-blown smoke constantly changing, color varies greatly from one combustion chamber another. Therefore, widely used sensor-based fire detection systems have disadvantages untimely high false rate in middle an open environment. Deep learning-based object can recognize objects form...
Traffic signs detection and recognition is an essential challenging task for driverless cars. However, the of traffic in most scenarios belongs to small target detection, existing object methods show poor performance these cases, which increases difficulty detection. To further improve accuracy signs, this paper proposed optimization strategy based on YOLOv4 network. Firstly, improved triplet attention mechanism was added backbone It combined with optimized weights make network focus more...
Lane line detection is a fundamental and critical task for geographic information perception of driverless advanced assisted driving. However, the traditional lane method relies on manual adjustment parameters, has poor universality, heavy workload, robustness. Most deep learning-based methods make it difficult to effectively balance accuracy efficiency. To improve comprehensive ability in natural traffic environment, algorithm based mixed-attention mechanism residual network (ResNet) row...
Semantic segmentation of high-resolution remote sensing images plays an important role in the community. However, many indistinguishable objects are prevalent within urban images, and some belonging to same class different that do not belong similar. These tricky make exhibit low-interclass variance high-intraclass variance, which significantly limits performance. Therefore, a fresh insight was presented alleviate this issue by incorporating fuzzy pattern recognition method deep-learning...
Abstract 3D object detection is a critical task in the fields of virtual reality and autonomous driving. Given that each sensor has its own strengths limitations, multi-sensor-based gained popularity. However, most existing methods extract high-level image semantic features fuse them with point cloud features, focusing solely on consistent information from both sensors while ignoring their complementary information. In this paper, we present novel two-stage multi-sensor deep neural network,...
In order to solve the problems of traffic object detection, fuzzification, and simplification in real environment, an automatic detection classification algorithm for roads, vehicles, pedestrians with multiple objects under same framework is proposed. We construct final V view through a considerate U-V method, which determines location horizon initial contour road. Road results are obtained error label reclassification, omitting point reassignment, so an. propose peripheral envelope...
Shadows and normal light illumination road non-road areas are two pairs of contradictory symmetrical individuals. To achieve accurate detection, it is necessary to remove interference caused by uneven illumination, such as shadows. This paper proposes a detection algorithm based on learning illumination-independent image solve the following problems: First, most methods sensitive variation illumination. Second, with traditional invariability, difficult determine calibration angle camera...
The rapid development of the economy has promoted growth freight transportation. truck service areas on expressways, as main places for drivers to rest, play an important role in ensuring driving safety trucks. If these are constructed densely or provide a plentiful supply parking areas, they costly construct. However, if distance between two adjacent is very large number spaces small, will fail meet needs drivers. In this situation, continuous working time be longer, and likely cause driver...
Surveillance video has been widely used in business, security, search, and other fields. Identifying locating specific pedestrians surveillance an important application value criminal investigation, search rescue, etc. However, the requirements for real-time capturing accuracy are high these applications. It is essential to build a complete smooth system combine pedestrian detection, tracking re-identification achieve goal of maximizing efficiency by balancing capture accuracy. This paper...
Unmanned Aerial Vehicle (UAV) aerial sensors are an important means of collecting ground image data. Through the road segmentation and vehicle detection drivable areas in UAV images, they can be applied to monitoring roads, traffic flow detection, management, etc. As well, integrated with intelligent transportation systems support related work departments. Existing algorithms only realize a single task, while requires simultaneous processing multiple tasks, which cannot meet complex...