Real-time Hand Pose Recognition Using Low-Resolution Depth Images
Computer Vision and Pattern Recognition, 2006 IEEE Computer Society Conference(2006)
摘要
Gesture recognition methods based on intensity or color images often suffer from low efficiency and lack of robustness. In this paper, we employ a new laser-based camera that produces reliable low-resolution depth images at video rates. By decomposing and recognizing hand poses as finger states (finger poses and finger inter-relations), we achieve robust hand pose recognition in real-time (30 frames/second).
更多查看译文
关键词
color image,real-time hand pose recognition,finger inter-relations,low-resolution depth images,new laser-based camera,video rate,gesture recognition method,low efficiency,reliable low-resolution depth image,robust hand,finger state,computer graphics,feature extraction,computer vision,gesture recognition,training data,robustness,image recognition,real time,color,low resolution
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要