Sleep monitoring via depth video recording and analysis

Cheng Yang, G. Cheung, K. Chan, V. Stankovic

Research output: Contribution to journalConference Contribution

9 Citations (Scopus)

Abstract

Quality of sleep greatly affects a person's physiological well-being. Traditional sleep monitoring systems are expensive in cost and intrusive enough that they disturb natural sleep of clinical patients. In this paper, we propose an inexpensive non-intrusive sleep monitoring system using recorded depth video only. In particular, we propose a two-part solution composed of depth video compression and analysis. For acquisition and compression, we first propose an alternating-frame video recording scheme, so that different 8 of the 11 bits in MS Kinect captured depth images are extracted at different instants for efficient encoding using H.264 video codec. At decoder, the uncoded 3 bits in each frame can be recovered accurately via a block-based search procedure. For analysis, we estimate parameters of our proposed dual-ellipse model in each depth image. Sleep events are then detected via a support vector machine trained on statistics of estimated ellipse model parameters over time. Experimental results show first that our depth video compression scheme outperforms a competing scheme that records only the eight most significant bits in PSNR in mid- to high-bitrate regions. Further, we show also that our monitoring can detect critical sleep events such as hypopnoea using our trained SVM with very high success rate.

Fingerprint

Video recording
Monitoring
Image compression
Support vector machines
Sleep
Statistics
Costs

Keywords

  • bioelectric potentials
  • biomedical optical imaging
  • error statistics
  • medical image processing
  • image coding
  • neurophysiology
  • support vector machines

Cite this

@article{e41684703e7a41af9fe1b97b5213ece8,
title = "Sleep monitoring via depth video recording and analysis",
abstract = "Quality of sleep greatly affects a person's physiological well-being. Traditional sleep monitoring systems are expensive in cost and intrusive enough that they disturb natural sleep of clinical patients. In this paper, we propose an inexpensive non-intrusive sleep monitoring system using recorded depth video only. In particular, we propose a two-part solution composed of depth video compression and analysis. For acquisition and compression, we first propose an alternating-frame video recording scheme, so that different 8 of the 11 bits in MS Kinect captured depth images are extracted at different instants for efficient encoding using H.264 video codec. At decoder, the uncoded 3 bits in each frame can be recovered accurately via a block-based search procedure. For analysis, we estimate parameters of our proposed dual-ellipse model in each depth image. Sleep events are then detected via a support vector machine trained on statistics of estimated ellipse model parameters over time. Experimental results show first that our depth video compression scheme outperforms a competing scheme that records only the eight most significant bits in PSNR in mid- to high-bitrate regions. Further, we show also that our monitoring can detect critical sleep events such as hypopnoea using our trained SVM with very high success rate.",
keywords = "bioelectric potentials, biomedical optical imaging, error statistics, medical image processing, image coding, neurophysiology, support vector machines",
author = "Cheng Yang and G. Cheung and K. Chan and V. Stankovic",
year = "2014",
month = "7",
day = "18",
doi = "10.1109/ICMEW.2014.6890645",
language = "English",
journal = "IEEE International Conference on Multimedia and Expo Workshops (ICMEW)",
issn = "1945-7871",
publisher = "IEEE",

}

Sleep monitoring via depth video recording and analysis. / Yang, Cheng; Cheung, G.; Chan, K.; Stankovic, V.

In: IEEE International Conference on Multimedia and Expo Workshops (ICMEW), 18.07.2014.

Research output: Contribution to journalConference Contribution

TY - JOUR

T1 - Sleep monitoring via depth video recording and analysis

AU - Yang, Cheng

AU - Cheung, G.

AU - Chan, K.

AU - Stankovic, V.

PY - 2014/7/18

Y1 - 2014/7/18

N2 - Quality of sleep greatly affects a person's physiological well-being. Traditional sleep monitoring systems are expensive in cost and intrusive enough that they disturb natural sleep of clinical patients. In this paper, we propose an inexpensive non-intrusive sleep monitoring system using recorded depth video only. In particular, we propose a two-part solution composed of depth video compression and analysis. For acquisition and compression, we first propose an alternating-frame video recording scheme, so that different 8 of the 11 bits in MS Kinect captured depth images are extracted at different instants for efficient encoding using H.264 video codec. At decoder, the uncoded 3 bits in each frame can be recovered accurately via a block-based search procedure. For analysis, we estimate parameters of our proposed dual-ellipse model in each depth image. Sleep events are then detected via a support vector machine trained on statistics of estimated ellipse model parameters over time. Experimental results show first that our depth video compression scheme outperforms a competing scheme that records only the eight most significant bits in PSNR in mid- to high-bitrate regions. Further, we show also that our monitoring can detect critical sleep events such as hypopnoea using our trained SVM with very high success rate.

AB - Quality of sleep greatly affects a person's physiological well-being. Traditional sleep monitoring systems are expensive in cost and intrusive enough that they disturb natural sleep of clinical patients. In this paper, we propose an inexpensive non-intrusive sleep monitoring system using recorded depth video only. In particular, we propose a two-part solution composed of depth video compression and analysis. For acquisition and compression, we first propose an alternating-frame video recording scheme, so that different 8 of the 11 bits in MS Kinect captured depth images are extracted at different instants for efficient encoding using H.264 video codec. At decoder, the uncoded 3 bits in each frame can be recovered accurately via a block-based search procedure. For analysis, we estimate parameters of our proposed dual-ellipse model in each depth image. Sleep events are then detected via a support vector machine trained on statistics of estimated ellipse model parameters over time. Experimental results show first that our depth video compression scheme outperforms a competing scheme that records only the eight most significant bits in PSNR in mid- to high-bitrate regions. Further, we show also that our monitoring can detect critical sleep events such as hypopnoea using our trained SVM with very high success rate.

KW - bioelectric potentials

KW - biomedical optical imaging

KW - error statistics

KW - medical image processing

KW - image coding

KW - neurophysiology

KW - support vector machines

UR - http://www.hot3d.org/

U2 - 10.1109/ICMEW.2014.6890645

DO - 10.1109/ICMEW.2014.6890645

M3 - Conference Contribution

JO - IEEE International Conference on Multimedia and Expo Workshops (ICMEW)

T2 - IEEE International Conference on Multimedia and Expo Workshops (ICMEW)

JF - IEEE International Conference on Multimedia and Expo Workshops (ICMEW)

SN - 1945-7871

ER -