Title :
On-bed sleep posture recognition based on body-earth mover´s distance
Author :
Xiaowei Xu;Feng Lin;Aosen Wang;Chen Song;Yu Hu;Wenyao Xu
Author_Institution :
School of Optical and Electronic Information, HUST, Wuhan, China
Abstract :
Sleep quality is highly correlated with many chronic diseases, such as cardiovascular disease and sleep disorders. Sleep posture is a key component in sleep quality assessment. Currently, body pressure analysis has been a popular approach for sleep posture recognition [1]. In this paper, we propose a novel distance metric for sleep posture dissimilarity measure, called Body-Earth Mover´s Distance (BEMD). BEMD can effectively recognize the difference of pressure distributions by projecting the pressure distribution to the horizontal and vertical directions. A pilot study is performed with 14 persons for six different postures. The experimental results show that our proposed BEMD can achieve a high accuracy of 90.78%, which outperforms the previous method [1] with an improvement of 9.1%.
Keywords :
"Sleep apnea","Histograms","Image segmentation","Fetus","Diseases","Measurement","Sensors"
Conference_Titel :
Biomedical Circuits and Systems Conference (BioCAS), 2015 IEEE
DOI :
10.1109/BioCAS.2015.7348281