A5102937604- Chaos-based Image/Signal Encryption
- Advanced Steganography and Watermarking Techniques
- Gaze Tracking and Assistive Technology
- Cellular Automata and Applications
- Retinal Imaging and Analysis
- Metal Extraction and Bioleaching
- Glaucoma and retinal disorders
- Digital Media Forensic Detection
- Digital Rights Management and Security
- Chaos control and synchronization
- Robotics and Sensor-Based Localization
- Vehicle License Plate Recognition
- Visual Attention and Saliency Detection
- Graph Theory and Algorithms
- Robotics and Automated Systems
- Optimization and Search Problems
- Computer Graphics and Visualization Techniques
- Recycling and utilization of industrial and municipal waste in materials production
- Image Processing and 3D Reconstruction
- Robotic Path Planning Algorithms
- Concrete and Cement Materials Research
China Jiliang University
2024-2025
Fuzhou University
2025
Northwestern Polytechnical University
2020-2024
Civil Aviation University of China
2020
Zhuhai Institute of Advanced Technology
2019
Beijing Institute of Technology
2019
Institute of Computing Technology
2016
Chinese Academy of Sciences
2016
North China University of Technology
2006-2007
In this paper, a color image encryption method based on finger vein key and off-axis digital holography with phase-modulated reference light is proposed. the process, firstly channel separation operation performed plaintext image, "red", "green" "blue" channels grayscale data of are obtained respectively. Subsequently, quantum matrix user generated through generation program used as scrambling index key, mask phase keys in next steps. The each original then encrypted using DNA coding...
A fast deblurring network, based on a high-performance convolutional network and pixel volume, is proposed to address the limitations of existing video algorithms, which often overly emphasize inter-frame information, leading high algorithmic complexity. First, networks are utilized prune thereby reducing both number model parameters computational To increased complexity resulting from extensive use traditional two-dimensional layers, depthwise over-parameterized convolutions employed...
Gaze estimation is an important fundamental task in computer vision and medical research. Existing works have explored various effective paradigms modules for precisely predicting eye gazes. However, the uncertainty gaze estimation, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">e.g</i> ., input annotation uncertainty, been neglected previous models use a deterministic function to estimate gaze, which cannot reflect actual situation...
Saccadic scanpath, a data representation of human visual behavior, has received broad interest in multiple domains. Scanpath is complex eye-tracking modality that includes the sequences fixation positions and duration, coupled with image information. However, previous methods usually face spatial misalignment problem features loss critical temporal (including correlation duration). In this study, we propose Transformer-based scanpath model, SpFormer, to alleviate these problems. First,...
In this paper, a three-dimensional (3D) image encryption method is proposed based on structured light illumination and an iterative layer-oriented angular-spectrum algorithm, where the original 3D plaintext encrypted into phase-only hologram ciphertext (POHC). The generated by using phase mask (SPM), optical parameters in SPM all serve as supplementary keys for increasing types quantities of security keys, expanding key space enhancing level security. Moreover, introduction also enhances...
Based on a class of orthogonal functions, novel digital watermarking method is proposed. This can be used in the anti counterfeit commercial bills. The experiment results show that algorithm more robust than traditional watermark algorithm. property uniformity corresponding Fibonacci transformation, proposed has following advantages: (1) encoding and decoding are very simple they applied real-time situations, (2) endure severe attacks such as extreme noise levels, high loss its data or...
Semitensor product theory can deal with matrices multiplication different numbers of columns and rows. Therefore, a new chaotic system for high dimensions be created by employing semitensor systems dimensions, so that more channels selected encryption. This paper proposes generated applied on Qi Lorenz systems. The corresponding dynamic characteristics the are discussed in this to verify existences attractors. detailed algorithms illustrated paper. FPGA hardware encryption implementations...
Based on the V- descriptors, a novel watermarking algorithm to protect copyright of vector images is proposed in this paper. Watermarks are embedded image's contours. Experiments show that robust against some attacks such as translation, scaling, rotation and local revision.
The based on the background of anti-counterfeit commercial bills, a novel digital watermarking method is proposed in this paper. algorithm class orthogonal function systems—V systems. experiment results show that has better robust than traditional watermark algorithm. Based property uniformity corresponding Fibonacci transformation, following advantages: (1) Encoding and decoding are very simple they can be applied real-time situations. (2) endure severe attacks such as extreme noise levels,...
This paper mainly studies the construction of two-dimensional map by lidar, mobile mode McCarnum wheel chassis, attitude and odometer information Beidou navigation system in ROS (Robot Operating System), real-time map, image generated simulation environment.