Body-Earth Mover's Distance: A Matching-Based Approach for Sleep Posture Recognition

Xiaowei Xu; Feng Lin; Aosen Wang; Yu Hu; Ming Chun Huang; Wenyao Xu

doi:10.1109/TBCAS.2016.2543686

Body-Earth Mover's Distance: A Matching-Based Approach for Sleep Posture Recognition

Xiaowei Xu
, Feng Lin
, Aosen Wang
, Yu Hu
, Ming Chun Huang
, Wenyao Xu

Department of Computer Science and Engineering

Huazhong University of Science and Technology
SUNY Buffalo
Case Western Reserve University

Research output: Contribution to journal › Article › peer-review

48 Scopus citations

Abstract

Sleep posture is a key component in sleep quality assessment and pressure ulcer prevention. Currently, body pressure analysis has been a popular method for sleep posture recognition. In this paper, a matching-based approach, Body-Earth Mover's Distance (BEMD), for sleep posture recognition is proposed. BEMD treats pressure images as weighted 2D shapes, and combines EMD and Euclidean distance for similarity measure. Compared with existing work, sleep posture recognition is achieved with posture similarity rather than multiple features for specific postures. A pilot study is performed with 14 persons for six different postures. The experimental results show that the proposed BEMD can achieve 91.21% accuracy, which outperforms the previous method with an improvement of 8.01%.

Original language	English
Article number	7524686
Pages (from-to)	1023-1035
Number of pages	13
Journal	IEEE Transactions on Biomedical Circuits and Systems
Volume	10
Issue number	5
DOIs	https://doi.org/10.1109/TBCAS.2016.2543686
State	Published - Oct 2016

Keywords

Earth Mover's Distance
pressure image
shape descriptor
sleep posture analysis

Access to Document

10.1109/TBCAS.2016.2543686

Cite this

@article{8a9f8f1c757445a3a79629836db26fcb,

title = "Body-Earth Mover's Distance: A Matching-Based Approach for Sleep Posture Recognition",

abstract = "Sleep posture is a key component in sleep quality assessment and pressure ulcer prevention. Currently, body pressure analysis has been a popular method for sleep posture recognition. In this paper, a matching-based approach, Body-Earth Mover's Distance (BEMD), for sleep posture recognition is proposed. BEMD treats pressure images as weighted 2D shapes, and combines EMD and Euclidean distance for similarity measure. Compared with existing work, sleep posture recognition is achieved with posture similarity rather than multiple features for specific postures. A pilot study is performed with 14 persons for six different postures. The experimental results show that the proposed BEMD can achieve 91.21\% accuracy, which outperforms the previous method with an improvement of 8.01\%.",

keywords = "Earth Mover's Distance, pressure image, shape descriptor, sleep posture analysis",

author = "Xiaowei Xu and Feng Lin and Aosen Wang and Yu Hu and Huang, \{Ming Chun\} and Wenyao Xu",

note = "Publisher Copyright: {\textcopyright} 2016 IEEE.",

year = "2016",

month = oct,

doi = "10.1109/TBCAS.2016.2543686",

language = "English",

volume = "10",

pages = "1023--1035",

journal = "IEEE Transactions on Biomedical Circuits and Systems",

issn = "1932-4545",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "5",

}

TY - JOUR

T1 - Body-Earth Mover's Distance

T2 - A Matching-Based Approach for Sleep Posture Recognition

AU - Xu, Xiaowei

AU - Lin, Feng

AU - Wang, Aosen

AU - Hu, Yu

AU - Huang, Ming Chun

AU - Xu, Wenyao

PY - 2016/10

Y1 - 2016/10

N2 - Sleep posture is a key component in sleep quality assessment and pressure ulcer prevention. Currently, body pressure analysis has been a popular method for sleep posture recognition. In this paper, a matching-based approach, Body-Earth Mover's Distance (BEMD), for sleep posture recognition is proposed. BEMD treats pressure images as weighted 2D shapes, and combines EMD and Euclidean distance for similarity measure. Compared with existing work, sleep posture recognition is achieved with posture similarity rather than multiple features for specific postures. A pilot study is performed with 14 persons for six different postures. The experimental results show that the proposed BEMD can achieve 91.21% accuracy, which outperforms the previous method with an improvement of 8.01%.

AB - Sleep posture is a key component in sleep quality assessment and pressure ulcer prevention. Currently, body pressure analysis has been a popular method for sleep posture recognition. In this paper, a matching-based approach, Body-Earth Mover's Distance (BEMD), for sleep posture recognition is proposed. BEMD treats pressure images as weighted 2D shapes, and combines EMD and Euclidean distance for similarity measure. Compared with existing work, sleep posture recognition is achieved with posture similarity rather than multiple features for specific postures. A pilot study is performed with 14 persons for six different postures. The experimental results show that the proposed BEMD can achieve 91.21% accuracy, which outperforms the previous method with an improvement of 8.01%.

KW - Earth Mover's Distance

KW - pressure image

KW - shape descriptor

KW - sleep posture analysis

UR - https://www.scopus.com/pages/publications/84979943753

U2 - 10.1109/TBCAS.2016.2543686

DO - 10.1109/TBCAS.2016.2543686

M3 - Article

C2 - 27483475

AN - SCOPUS:84979943753

SN - 1932-4545

VL - 10

SP - 1023

EP - 1035

JO - IEEE Transactions on Biomedical Circuits and Systems

JF - IEEE Transactions on Biomedical Circuits and Systems

IS - 5

M1 - 7524686

ER -

Body-Earth Mover's Distance: A Matching-Based Approach for Sleep Posture Recognition

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this