A5058067357- 3D Shape Modeling and Analysis
- Human Pose and Action Recognition
- Face recognition and analysis
- 3D Surveying and Cultural Heritage
- Advanced Image and Video Retrieval Techniques
- Image Processing and 3D Reconstruction
- Biometric Identification and Security
- Advanced Neural Network Applications
- Gait Recognition and Analysis
- Video Surveillance and Tracking Methods
- Advanced Vision and Imaging
- Hand Gesture Recognition Systems
Shenzhen University
2024
Guangdong University of Technology
2020
3D face recognition performs better than 2D recognition, in terms of robustness against lighting and digital attacks. Compared to data, the rich geometric information data could be very useful improve performance recognition. This paper introduces PointSurFace, a surface feature extraction integrated model tailored for point cloud based PointSurFace adopts specially designed module, which explores about position normals explicitly extract local discriminative features from faces....
Recognizing the accurate position of each joint point human body helps computer to understand rich motion information. precise in 3D space In order avoid defects binocular vision combined with neural network method, we have designed an optimization algorithm for pose recognition technology under vision. The uses HOG and LBP feature operators perform matching on corresponding areas left right images, seeks optimal pixel joints points. And search window can adjust itself best size. This...
This paper proposes a human motion capture system based on multi-binocular vision and image recognition. The which relies the principle of binocular imaging stereo matching same name points joints in left right views to obtain 3D coordinates joint real time, uses multiple cameras placed at different angles for simultaneous shooting, expanding integrity original system. At detection model abnormal estimation under is established, so that can more accurately identify due occlusion. Finally,...