作者: Xiaowei Xu , Feng Lin , Aosen Wang , Yu Hu , Ming-Chun Huang
DOI: 10.1109/TBCAS.2016.2543686
关键词: Similarity measure 、 Computer vision 、 Polysomnography 、 Pattern recognition (psychology) 、 Similarity (network science) 、 Matching (statistics) 、 Euclidean distance 、 Artificial intelligence 、 Work (physics) 、 Earth mover's distance 、 Engineering
摘要: Sleep posture is a key component in sleep quality assessment and pressure ulcer prevention. Currently, body analysis has been popular method for recognition. In this paper, matching-based approach, Body-Earth Mover’s Distance (BEMD), recognition proposed. BEMD treats images as weighted 2D shapes, combines EMD Euclidean distance similarity measure. Compared with existing work, achieved rather than multiple features specific postures. A pilot study performed 14 persons six different The experimental results show that the proposed can achieve 91.21% accuracy, which outperforms previous an improvement of 8.01%.