A Novel Framework for Remote Photoplethysmography Pulse Extraction on Compressed Videos

Changchen Zhao, Chun-Liang Lin, Weihai Chen, Zhengguo Li; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2018, pp. 1299-1308

Abstract


Remote photoplethysmography (rPPG) has recently attracted much attention due to its non-contact measurement convenience and great potential in health care and computer vision applications. However, almost all the existing rPPG methods are based on uncompressed video data, which greatly limits its application to the scenarios that require long-distance video transmission. This paper proposes a novel framework as a first attempt to address the rPPG pulse extraction in presence of video compression artifacts. Based on the analysis of the impact of various compression methods on rPPG measurements, the problem is cast as single-channel signal separation. The framework consists of three major steps to extract the pulse waveform and heart rate by exploiting frequency structure of the rPPG signal. A benchmark dataset which contains stationary and motion videos has been built. The results show that the proposed algorithm significantly improves the SNR and heart rate precision of state-of-the-art rPPG algorithms on stationary videos and has a positive effect on motion videos at low bitrates.

Related Material


[pdf]
[bibtex]
@InProceedings{Zhao_2018_CVPR_Workshops,
author = {Zhao, Changchen and Lin, Chun-Liang and Chen, Weihai and Li, Zhengguo},
title = {A Novel Framework for Remote Photoplethysmography Pulse Extraction on Compressed Videos},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2018}
}