Fig 1.
Face video HR estimation framework.
Fig 2.
Face and skin detection results.
A:The result of face detection. B:The cropped face. C∼F: The skin detection results of RGB, YCrCb, YCrCb+OTSU, RGB-H.
Fig 3.
The IMFs and the residual of the filtered signal.
Fig 4.
Frequency domain analysis of the decomposed IMF components.
Fig 5.
The waveforms of Green channel and after pretreatment.
Fig 6.
Original signal and its first five components.
Fig 7.
Proportion curves of singular values.
Fig 8.
The autocorrelation of HR-related signal and noise signal.
A: The HR-related signal(left), the noise signal(right). B: The autocorrelation of HR-related signal(left), The autocorrelation of noise signal(right).
Fig 9.
Autocorrelation coefficient Pi(k) and maximum autocorrelation coefficient ρi of component i.
Fig 10.
Screening of the first s maximum autocorrelation coefficients.
Fig 11.
Waveform comparison of filtered G-channel, EEMD and LA-SSA.
Fig 12.
Comparison of LA-SSA result with reference PPG signal.
A: Waveform comparison. B: Spectrum comparison.
Table 1.
EALM algorithm.
Table 2.
Effect of different threshold on HR estimation accuracy.
Table 3.
Performance comparison of LA-SSA HR estimation using different input signals.
Fig 13.
The comparison of HR values obtained by the proposed method and the contact measurement method.
A: Linear regression diagram. B: Bland-Altman diagram.
Table 4.
Performance for different noise level of LA-SSA HR estimation.
Table 5.
Performance of different methods for HR estimation on database UBFC-RPPG.
Table 6.
Performance of different methods for HR estimation on database PURE.