def derivative_filter(ecg):
# Make impulse response
h = [-1, -2, 0, 2, 1]
h = [x / 8 for x in h]
# Apply filter
ecg4 = np.convolve(ecg, h)
ecg4 = np.roll(ecg4, -6)
return normalizer.max_normalization(ecg4)
def moving_window_integration(ecg):
# Make impulse response
h = np.ones(31)
h = np.array([x / 31 for x in h])
# Apply filter
ecg6 = np.convolve(ecg, h)
ecg6 = np.roll(ecg6, -15)
return normalizer.max_normalization(ecg6)
def low_pass_filtering(ecg):
# LPF (1-z^-6)^2/(1-z^-1)^2
b = [1, 0, 0, 0, 0, 0, -2, 0, 0, 0, 0, 0, 1]
a = [1, -2, 1]
# transfer function of LPF
h_LP = lfilter(b, a, np.append([1], np.zeros(12)))
ecg2 = np.convolve(ecg, h_LP)
# cancel delay
ecg2 = np.roll(ecg2, -6)
return normalizer.max_normalization(ecg2)
def high_pass_filtering(ecg):
# HPF = Allpass-(Lowpass) = z^-16-[(1-z^32)/(1-z^-1)]
b = [
-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 32, -32, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1
]
a = [1, -1]
# impulse response iof HPF
h_HP = lfilter(b, a, np.append([1], np.zeros(32)))
ecg3 = np.convolve(ecg, h_HP)
# cancel delay
ecg3 = np.roll(ecg3, -16)
return normalizer.max_normalization(ecg3)
def squaring(ecg):
ecg5 = np.square(ecg)
return normalizer.max_normalization(ecg5)