Resumé

Automatic analysis of body movement to identify physical activity of patients who are at bed rest is crucial for treatment or rehabilitation purposes. Existing methods of physical activity analysis mostly focused on the detection of primitive motion/non-motion states in unimodal video data captured by either RGB or depth or thermal sensor. In this paper, we propose a multimodal visionbased approach to classify body motion of a person lying on a bed. We mimicked a real scenario of ’patient on bed’ by recording multimodal video data from healthy volunteers in a hospital room in a neurorehabilitation center. We first defined a taxonomy of possible physical activities based on observations of patients with acquired brain injuries. We then investigated different motion analysis and machine learning approaches to classify physical activities automatically. A multimodal database including RGB, depth and thermal videos was collected and annotated with eight predefined physical activities. Experimental results show that we can achieve moderately high accuracy (77.68%) to classify physical activities by tracking the body motion using an optical flow-based approach. To the best of our knowledge this is the first multimodal RGBDT video analysis for such application.
OriginalsprogEngelsk
TitelAdvances in Visual Computing : 13th International Symposium, ISVC 2018, Las Vegas, NV, USA, November 19 – 21, 2018, Proceedings
Antal sider12
ForlagSpringer
Publikationsdato2018
Sider552-564
ISBN (Trykt)978-3-030-03800-7
ISBN (Elektronisk)978-3-030-03801-4
DOI
StatusUdgivet - 2018
BegivenhedInternational Symposium on Visual Computing, ISVC 2018 - Las Vegas, USA
Varighed: 19 nov. 201821 nov. 2018
Konferencens nummer: 13

Konference

KonferenceInternational Symposium on Visual Computing, ISVC 2018
Nummer13
LandUSA
ByLas Vegas
Periode19/11/201821/11/2018
NavnLecture Notes in Computer Science
Vol/bind11241
ISSN0302-9743

Fingerprint

Optical flows
Taxonomies
Patient rehabilitation
Learning systems
Brain
Sensors
Hot Temperature
Motion analysis

Citer dette

Haque, M. A., Kjeldsen, S. S., Arguissain, F. G., Brunner, I., Nasrollahi, K., Andersen, O. K., ... Jørgensen, A. (2018). Patient’s Body Motion Study using Multimodal RGBDT Videos. I Advances in Visual Computing: 13th International Symposium, ISVC 2018, Las Vegas, NV, USA, November 19 – 21, 2018, Proceedings (s. 552-564). Springer. Lecture Notes in Computer Science, Bind. 11241 https://doi.org/10.1007/978-3-030-03801-4_48
Haque, Mohammad Ahsanul ; Kjeldsen, Simon S. ; Arguissain, Federico G. ; Brunner, Iris ; Nasrollahi, Kamal ; Andersen, Ole Kæseler ; Nielsen, Jørgen F. ; Moeslund, Thomas B. ; Jørgensen, Anders. / Patient’s Body Motion Study using Multimodal RGBDT Videos. Advances in Visual Computing: 13th International Symposium, ISVC 2018, Las Vegas, NV, USA, November 19 – 21, 2018, Proceedings. Springer, 2018. s. 552-564 (Lecture Notes in Computer Science, Bind 11241).
@inproceedings{946a770ab8ba425aa54ff6534b78fb15,
title = "Patient’s Body Motion Study using Multimodal RGBDT Videos",
abstract = "Automatic analysis of body movement to identify physical activity of patients who are at bed rest is crucial for treatment or rehabilitation purposes. Existing methods of physical activity analysis mostly focused on the detection of primitive motion/non-motion states in unimodal video data captured by either RGB or depth or thermal sensor. In this paper, we propose a multimodal visionbased approach to classify body motion of a person lying on a bed. We mimicked a real scenario of ’patient on bed’ by recording multimodal video data from healthy volunteers in a hospital room in a neurorehabilitation center. We first defined a taxonomy of possible physical activities based on observations of patients with acquired brain injuries. We then investigated different motion analysis and machine learning approaches to classify physical activities automatically. A multimodal database including RGB, depth and thermal videos was collected and annotated with eight predefined physical activities. Experimental results show that we can achieve moderately high accuracy (77.68{\%}) to classify physical activities by tracking the body motion using an optical flow-based approach. To the best of our knowledge this is the first multimodal RGBDT video analysis for such application.",
keywords = "Physical activity, Multimodal, RGBDT, Video, Rest activity, Patient, hospital, Bed, Bed Rest",
author = "Haque, {Mohammad Ahsanul} and Kjeldsen, {Simon S.} and Arguissain, {Federico G.} and Iris Brunner and Kamal Nasrollahi and Andersen, {Ole K{\ae}seler} and Nielsen, {J{\o}rgen F.} and Moeslund, {Thomas B.} and Anders J{\o}rgensen",
year = "2018",
doi = "10.1007/978-3-030-03801-4_48",
language = "English",
isbn = "978-3-030-03800-7",
pages = "552--564",
booktitle = "Advances in Visual Computing",
publisher = "Springer",
address = "Germany",

}

Haque, MA, Kjeldsen, SS, Arguissain, FG, Brunner, I, Nasrollahi, K, Andersen, OK, Nielsen, JF, Moeslund, TB & Jørgensen, A 2018, Patient’s Body Motion Study using Multimodal RGBDT Videos. i Advances in Visual Computing: 13th International Symposium, ISVC 2018, Las Vegas, NV, USA, November 19 – 21, 2018, Proceedings. Springer, Lecture Notes in Computer Science, bind 11241, s. 552-564, Las Vegas, USA, 19/11/2018. https://doi.org/10.1007/978-3-030-03801-4_48

Patient’s Body Motion Study using Multimodal RGBDT Videos. / Haque, Mohammad Ahsanul; Kjeldsen, Simon S.; Arguissain, Federico G.; Brunner, Iris; Nasrollahi, Kamal; Andersen, Ole Kæseler; Nielsen, Jørgen F. ; Moeslund, Thomas B.; Jørgensen, Anders.

Advances in Visual Computing: 13th International Symposium, ISVC 2018, Las Vegas, NV, USA, November 19 – 21, 2018, Proceedings. Springer, 2018. s. 552-564 (Lecture Notes in Computer Science, Bind 11241).

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

TY - GEN

T1 - Patient’s Body Motion Study using Multimodal RGBDT Videos

AU - Haque, Mohammad Ahsanul

AU - Kjeldsen, Simon S.

AU - Arguissain, Federico G.

AU - Brunner, Iris

AU - Nasrollahi, Kamal

AU - Andersen, Ole Kæseler

AU - Nielsen, Jørgen F.

AU - Moeslund, Thomas B.

AU - Jørgensen, Anders

PY - 2018

Y1 - 2018

N2 - Automatic analysis of body movement to identify physical activity of patients who are at bed rest is crucial for treatment or rehabilitation purposes. Existing methods of physical activity analysis mostly focused on the detection of primitive motion/non-motion states in unimodal video data captured by either RGB or depth or thermal sensor. In this paper, we propose a multimodal visionbased approach to classify body motion of a person lying on a bed. We mimicked a real scenario of ’patient on bed’ by recording multimodal video data from healthy volunteers in a hospital room in a neurorehabilitation center. We first defined a taxonomy of possible physical activities based on observations of patients with acquired brain injuries. We then investigated different motion analysis and machine learning approaches to classify physical activities automatically. A multimodal database including RGB, depth and thermal videos was collected and annotated with eight predefined physical activities. Experimental results show that we can achieve moderately high accuracy (77.68%) to classify physical activities by tracking the body motion using an optical flow-based approach. To the best of our knowledge this is the first multimodal RGBDT video analysis for such application.

AB - Automatic analysis of body movement to identify physical activity of patients who are at bed rest is crucial for treatment or rehabilitation purposes. Existing methods of physical activity analysis mostly focused on the detection of primitive motion/non-motion states in unimodal video data captured by either RGB or depth or thermal sensor. In this paper, we propose a multimodal visionbased approach to classify body motion of a person lying on a bed. We mimicked a real scenario of ’patient on bed’ by recording multimodal video data from healthy volunteers in a hospital room in a neurorehabilitation center. We first defined a taxonomy of possible physical activities based on observations of patients with acquired brain injuries. We then investigated different motion analysis and machine learning approaches to classify physical activities automatically. A multimodal database including RGB, depth and thermal videos was collected and annotated with eight predefined physical activities. Experimental results show that we can achieve moderately high accuracy (77.68%) to classify physical activities by tracking the body motion using an optical flow-based approach. To the best of our knowledge this is the first multimodal RGBDT video analysis for such application.

KW - Physical activity

KW - Multimodal

KW - RGBDT

KW - Video

KW - Rest activity

KW - Patient

KW - hospital

KW - Bed

KW - Bed Rest

U2 - 10.1007/978-3-030-03801-4_48

DO - 10.1007/978-3-030-03801-4_48

M3 - Article in proceeding

SN - 978-3-030-03800-7

SP - 552

EP - 564

BT - Advances in Visual Computing

PB - Springer

ER -

Haque MA, Kjeldsen SS, Arguissain FG, Brunner I, Nasrollahi K, Andersen OK et al. Patient’s Body Motion Study using Multimodal RGBDT Videos. I Advances in Visual Computing: 13th International Symposium, ISVC 2018, Las Vegas, NV, USA, November 19 – 21, 2018, Proceedings. Springer. 2018. s. 552-564. (Lecture Notes in Computer Science, Bind 11241). https://doi.org/10.1007/978-3-030-03801-4_48