A5100635903- Robotics and Sensor-Based Localization
- Infrared Target Detection Methodologies
- Advanced Vision and Imaging
- Advanced Measurement and Detection Methods
- 3D Surveying and Cultural Heritage
- Advanced Image and Video Retrieval Techniques
- Robotic Path Planning Algorithms
- Optical Systems and Laser Technology
- Image and Object Detection Techniques
- Infrastructure Maintenance and Monitoring
- Concrete Corrosion and Durability
- Inertial Sensor and Navigation
- Particle accelerators and beam dynamics
- Asphalt Pavement Performance Evaluation
- Industrial Vision Systems and Defect Detection
- Image Processing Techniques and Applications
- Human Pose and Action Recognition
- Magnetic confinement fusion research
- Calibration and Measurement Techniques
- Advanced Neural Network Applications
- Particle Accelerators and Free-Electron Lasers
- Smart Agriculture and AI
- Non-Destructive Testing Techniques
- Optical measurement and interference techniques
- Remote Sensing in Agriculture
Sun Yat-sen University
2018-2024
Ministry of Industry and Information Technology
2024
Nanjing Agricultural University
2024
Ministry of Agriculture and Rural Affairs
2024
The Synergetic Innovation Center for Advanced Materials
2024
Jiangsu University of Science and Technology
2023
Huazhong Agricultural University
2023
Nanjing University of Aeronautics and Astronautics
2016-2023
Beijing Building Construction Research Institute (China)
2022
Northwestern Polytechnical University
2021
Purpose This paper aims to reveal the corrosion mechanism and development regulation of marine engineering structural steel in environment provide constructive suggestions for immersed tunnel engineering. Design/methodology/approach In this study, steel’s behavior prediction were carried out under conditions no cathodic protection under-protection by artificially adding dissolved oxygen a simulated seawater solution as depolarizing agent. Findings Marine resources are rich China. With...
In the testing of chips, defect diagnostics in X-ray images packaging chips is mainly performed by humans, which time-consuming and inefficient. To overcome abovementioned problems, a novel intelligent system based on hybrid deep learning for chip was proposed. The consists four successive stages: image segmentation normalization, reconstruction detection, contour matching, qualification diagnosis. first stage used to localize external contours target remove extraneous backgrounds through...
Although space targets have different shapes, sizes and intensities, their distributions share certain commonalities. However, it is difficult to summarize a generalized distribution function for targets. Moreover, most of the existing methods based on deep learning are not suitable use directly because size cost manual labeling full image. In this paper, we proposed pattern target detection convolutional neural network (CNN) learn essential features from data. processing stage, background...
In the manufacturing of chips, accurate and effective detection internal bubble defects chips is essential to maintain product reliability. general, inspection performed manually by viewing X-ray images, which time-consuming less reliable. To solve above problems, an improved defect model YOLO-Xray based on YOLOv5 algorithm for chip images proposed. First, are preprocessed image segmentation construct dataset, namely, CXray. Then, in input stage, K-means++ used re-cluster CXray dataset...
We address the model-to-image registration problem with line features in following two ways. (a) present a robust solution to simultaneously recover camera pose and three-dimensional-to-two-dimensional correspondences. With weak priors, our approach progressively verifies guesses Kalman filter by using subset of recursively found match hypotheses. Experiments show method is occlusions clutter. (b) propose new feature based estimation algorithm, which iteratively optimizes objective function...
The registration of astronomical images, is one the key technologies to improve detection accuracy small and weak targets during image-based observation. error has a great influence on trace association targets. However, most existing methods for point-matching image transformation lack pertinence this actual scene. In study, we propose algorithm based geometric constraints homography, images. First, position changes in stars an caused by motion platform where camera had been stationed, were...
In space-based observations, there are large numbers of targets with size ranges and intensities in a given image. However, the current detection studies mostly emphasize saliency but neglect their distribution characteristics, which may result incorrect or missed results. this letter, an effective algorithm based on target characteristics is proposed to pursue good multitarget performance. First, space model studied obtain its characteristics. Next, operators applied reduce influence search...
With the increasing number of space debris in near-earth space, surveillance has become more crucial to prevent potential collisions. In an optical system, a multitude with varying sizes, speeds and intensities exist within images, which brings great difficulties for extraction targets. Current methodologies mainly emphasize local saliency enhancement source based on global statistical attributes, yet often overlook distribution similarity individuals how map them standardized space. To...
In recent years, the continuously increasing number of space objects has caught great attention and triggered need for surveillance prevention potential collisions. Moving may appear as streaks in optical images some situations. However, most them are usually unknown so that they lack sufficient prior information, which brings difficulty to detection. Motivated by fact have their unique image characteristics, this article presents a novel detection pattern streak-like targets single images....