TY - JOUR
T1 - Bounded kalman filter method for motion-robust, non-contact heart rate estimation
AU - Arul Prakash, Sakthi Kumar
AU - Tucker, Conrad S.
N1 - Funding Information:
National Center for Advancing Translational Sciences, National Institutes of Health (NIH), through Grant UL1 TR000127 and TR002014. National Science Foundation (NSF) NRI award #1527148 and NSF I/UCRC Center for Healthcare Organization Transformation (CHOT), NSF I/UCRC award #1624727. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NSF or NIH.
Publisher Copyright:
© 2018 Optical Society of America.
PY - 2018/2/1
Y1 - 2018/2/1
N2 - The authors of this work present a real-time measurement of heart rate across different lighting conditions and motion categories. This is an advancement over existing remote photo plethysmography (rPPG) methods that require a static, controlled environment for heart rate detection, making them impractical for real-world scenarios wherein a patient may be in motion, or remotely connected to a healthcare provider through telehealth technologies. The algorithm aims to minimize motion artifacts such as blurring and noise due to head movements (uniform, random) by employing i) a blur identification and denoising algorithm for each frame and ii) a bounded Kalman filter technique for motion estimation and feature tracking. A case study is presented that demonstrates the feasibility of the algorithm in non-contact estimation of the pulse rate of subjects performing everyday head and body movements. The method in this paper outperforms state of the art rPPG methods in heart rate detection, as revealed by the benchmarked results.
AB - The authors of this work present a real-time measurement of heart rate across different lighting conditions and motion categories. This is an advancement over existing remote photo plethysmography (rPPG) methods that require a static, controlled environment for heart rate detection, making them impractical for real-world scenarios wherein a patient may be in motion, or remotely connected to a healthcare provider through telehealth technologies. The algorithm aims to minimize motion artifacts such as blurring and noise due to head movements (uniform, random) by employing i) a blur identification and denoising algorithm for each frame and ii) a bounded Kalman filter technique for motion estimation and feature tracking. A case study is presented that demonstrates the feasibility of the algorithm in non-contact estimation of the pulse rate of subjects performing everyday head and body movements. The method in this paper outperforms state of the art rPPG methods in heart rate detection, as revealed by the benchmarked results.
UR - http://www.scopus.com/inward/record.url?scp=85041549143&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85041549143&partnerID=8YFLogxK
U2 - 10.1364/BOE.9.000873
DO - 10.1364/BOE.9.000873
M3 - Article
C2 - 29552419
AN - SCOPUS:85041549143
VL - 9
SP - 873
EP - 897
JO - Biomedical Optics Express
JF - Biomedical Optics Express
SN - 2156-7085
IS - 2
M1 - #312465
ER -