A5008194767- Advanced Image Processing Techniques
- Image and Signal Denoising Methods
- Retinal and Optic Conditions
- Retinal Imaging and Analysis
- Advanced Image and Video Retrieval Techniques
- Advanced Neural Network Applications
- Image Processing Techniques and Applications
- Robotics and Sensor-Based Localization
- Advanced Image Fusion Techniques
- Digital Imaging for Blood Diseases
- Computer Graphics and Visualization Techniques
- Advanced Optical Sensing Technologies
- Brain Tumor Detection and Classification
- Computational Geometry and Mesh Generation
- Infrared Target Detection Methodologies
- Advanced Vision and Imaging
- Video Surveillance and Tracking Methods
- Digital Media Forensic Detection
Chongqing University
2018-2023
China Mobile (China)
2015-2016
Central South University
2015-2016
Recent point-based differentiable rendering techniques have achieved significant success in high-fidelity reconstruction and fast rendering. However, due to the unstructured nature of representations, they are difficult apply modern graphics pipelines designed for structured meshes, as well a variety simulation editing algorithms that work with mesh representations. To this end, we propose StructuredField, novel representation achieves both geometric reconstructed object object. We employ...
An ensemble method based on supervised learning for segmenting the retinal vessels in color fundus images is proposed basis of previous work Zhu et al. For each pixel, a 36 dimensional feature vector extracted, including local features, morphological transformation with multi-scale and multi-orientation, divergence field which firstly used to extract image pixels. Then as input data set train weak classifiers by Classification regression tree (CART). Finally, an AdaBoost classifier...
ABSTRACT Accurate target detection in low‐light environments is crucial for unmanned aerial vehicles (UAVs) and autonomous driving applications. In this study, the authors introduce a real‐time multimodal fusion enhanced (RMF‐ED), novel framework designed to overcome limitations of detection. By leveraging complementary capabilities near‐infrared (NIR) cameras light ranging (LiDAR) sensors, RMF‐ED enhances performance. An advanced NIR generative adversarial network (NIR‐GAN) model was...
Object detection is one of the core technologies in aerial image processing and analysis. Although existing object methods based on deep learning have made some progress, there are still problems remained: (1) Most fail to simultaneously consider multi-scale multi-shape characteristics images, which may lead missing or false detections; (2) high precision generally requires a large complex network structure, usually makes it difficult achieve efficiency deploy resource-constrained devices...
Multi-scale object detection is a basic challenge in computer vision. Although many advanced methods based on convolutional neural networks have succeeded natural images, the progress aerial images has been relatively slow mainly due to considerably huge scale variations of objects and densely distributed small objects. In this paper, considering that semantic information may be weakened or even disappear deeper layers network, we propose new framework called Extended Feature Pyramid Network...
This paper propose an improved supervised method for retinal vessel segmentation based on Extreme Learning Machine (ELM). Firstly, a 36-D feature vector is extracted each pixel of the fund us image consisting local features, morphological features and divergence fields. Then matrix pixels training set using manual constructed as input ELM. Finally classifier obtained to segment vessels. The evaluated with DRIVE database average accuracy 0.9581. And running time greatly decreased by It...
This paper proposes a blur kernel estimation model based on combined constraints involving both image and for blind deblurring. We adopt L <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> regularization term constraining gradient dark channel of to protect strong edges suppress noise in image, use xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> as hybrid its preserve kernel's sparsity continuity respectively. In constraints, the...