Automatic RoI Detection for Camera-Based Pulse-Rate Measurement

Abstract

Abstract. Remote photoplethysmography (rPPG) enables contactless measurement of pulse-rate by detecting pulse-induced colour changes on human skin using a regular camera. Most of existing rPPG methods exploit the subject face as the Region of Interest (RoI) for pulse-rate measurement by automatic face detection. However, face detection is a suboptimal solution since (1) not all the subregions in a face contain the skin pixels where pulse-signal can be extracted, (2) it fails to locate the RoI in cases when the frontal face is invisible (e.g., side-view faces). In this paper, we present a novel automatic RoI detection method for camera-based pulse-rate measurement, which consists of three main steps: subregion tracking, feature extraction, and clustering of skin regions. To evaluate the robustness of the proposed method, 36 video recordings are made of 6 subjects with different skin-types performing 6 types of head motion. Experimental results show that for the video sequences containing subjects with brighter skin-types and modest body motions, the accuracy of the pulse-rates measured by our method (94%) is comparable to that obtained by a face detector (92%), while the average SNR is significantly improved from 5.8 dB to 8.6 dB.

Respiration Monitoring for Premature Neonates in NICU

“…Therefore, our method can improve the comfort level and convenience. A further benefit of using a camera is that it enables more measurements than contact-based bio-sensors, including physiological signals (e.g., breathing rate, heart rate, and blood oxygen saturation) [24,25] and contexture signals (e.g., body motion, activities, and facial expressions) [26][27][28][29]. This will enrich the functionality of a health monitoring system.…”

Feasibility of Specular Reflection Imaging for Extraction of Neck Vessel Pressure Waveforms

“…Spatial clustering was then performed on the global correlation map to identify disparate regions characterized by highly negative correlations and highly positive correlations. Spatial clustering has been performed in a similar application; specifically, ( Luijtelaar et al, 2014 ), performed a density-based clustering of signals represented in a feature space to segregate rPPG signals originating from tissue from noise signals originating from non-tissue regions in the video frame. The authors then performed a cluster-growing operation, such that signals originally classified as not being generated by skin may be re-added using information from the initial clustering.…”

Fully-Automatic Camera-Based Pulse-Oximetry During Sleep

“…A unique discriminative feature of skin pixels is the presence of cardiac-synchronous color variations induced by blood volume variations, which we refer to as 'living-skin', which was first exploited by Jeanne et al [9]. Elaborating on this idea, most methods in living-skin detection [9,5,11,21,12,1,24,25] use a common scheme consisting of three steps: (1) segmenting the video into spatio-temporal regions to extract locally independent rPPG-signals; (2) exploiting intrinsic properties of the pulse signal to differentiate pulse and noise from extracted rPPG signals; and (3) labeling the regions containing pulse as skin. In this scheme, the core function is step (2) that separates pulse and noise, which is also the key component to distinguish different methods in literature.…”

“…spectrum amplitude), as shown by the comparison in [24]. Van Luijtelaar et al [21] constructed a joint multi-dimensional feature space using different properties of pulse and skin, and applied a clustering method to find skin. A similarity-based living-skin detection method "Voxel-Pulse-Spectral" (VPS) has been proposed by Wang et al [24] to detect the regions sharing pulse similarities (e.g.…”